7103 research outputs found
Sort by
From welfare to work : the impact of the jobseeker-adviser relationship on objective employment outcomes
Get PDFRethinking ducted turbines : the fundamentals of aerodynamic performance and theory
Get PDFDucted turbines are designed to augment the flow through a rotor and consequently increase power extraction, with the aim of reducing the cost of wind energy. Despite many years of research, however, much uncertainty remains on a fundamental level: uncertainty that is not conducive to maximising performance or to commercial success. This work reduces the problem's complexity and improves understanding of the fundamentals by examining the underlying inviscid behaviour of ducted turbines, which are also known as diffuser augmented turbines. Numerical results show that the Betz limit does not apply, even using duct exit area, confirm the applicability of inviscid simulations to attached viscous flow, and clarify the influence of duct geometry. A comparison demonstrates that the diffuser conceptual model, which has dominated research thus far, is outperformed by an aerofoil conceptual model. The latter gives a closer match between intuition and actual performance, is easier to work with, and allows the influence of the rotor to be thought of as a change in the flow seen by the duct. It is therefore recommended as the standard for future studies. Theoretical examinations establish that invalid simplifying assumptions in existing theories leave the requirement for empirical parameters intact, and that velocity at the rotor may better fill the empirical parameter role than exit pressure or duct drag. A detailed derivation for the relationship between inviscid duct drag and augmentation is also described for the first time. An analysis suggests that ducts inherently reduce the optimum rotor loading in inviscid flow, with increases in rotor loading decreasing duct performance by reducing the effective duct wall angle and effective free stream velocity magnitude. Viscous effects may then increase the optimum, play a larger role than otherwise appears, and have greater potential for performance improvements than previously thought.Ducted turbines are designed to augment the flow through a rotor and consequently increase power extraction, with the aim of reducing the cost of wind energy. Despite many years of research, however, much uncertainty remains on a fundamental level: uncertainty that is not conducive to maximising performance or to commercial success. This work reduces the problem's complexity and improves understanding of the fundamentals by examining the underlying inviscid behaviour of ducted turbines, which are also known as diffuser augmented turbines. Numerical results show that the Betz limit does not apply, even using duct exit area, confirm the applicability of inviscid simulations to attached viscous flow, and clarify the influence of duct geometry. A comparison demonstrates that the diffuser conceptual model, which has dominated research thus far, is outperformed by an aerofoil conceptual model. The latter gives a closer match between intuition and actual performance, is easier to work with, and allows the influence of the rotor to be thought of as a change in the flow seen by the duct. It is therefore recommended as the standard for future studies. Theoretical examinations establish that invalid simplifying assumptions in existing theories leave the requirement for empirical parameters intact, and that velocity at the rotor may better fill the empirical parameter role than exit pressure or duct drag. A detailed derivation for the relationship between inviscid duct drag and augmentation is also described for the first time. An analysis suggests that ducts inherently reduce the optimum rotor loading in inviscid flow, with increases in rotor loading decreasing duct performance by reducing the effective duct wall angle and effective free stream velocity magnitude. Viscous effects may then increase the optimum, play a larger role than otherwise appears, and have greater potential for performance improvements than previously thought
Low-voltage organic transistors with high transconductance
Get PDFThis thesis presents the development of low-voltage organic thin-film transistors with high transconductance. This was achieved by employing ultra-thin bi-layer gate dielectric consisting of aluminium oxide (AlOx) and a self-assembled monolayer of octadecyl phosphonic acid (C18PA) and by increasing the channel width of the transistors through the implementation of the multi-finger source/drain contacts. The transistors based on dinaphtho[2,3-b:2',3'-f]thieno[3,2-b]thiophene (DNTT) exhibited low turn-on voltage and a.c. transconductance around 30 to 60 µS. Transistor amplifiers based on such transistors exhibited voltage gain approaching 10 V/V and a gain of about 2 V/V when the supply voltage was limited to 5 V. Next, a series of [n]phenacenes ([n] = 5, 6, or 7) was used for the first time in combination with the thin AlOx/C18PA dielectric bi-layer. Regardless of the substrate and the source-drain contact geometry, the field-effect mobility of such transistors was found to increase with increasing length of the conjugated [n]phenacene core, leading to the best performance for [7]phenacene with the largest average field-effect mobility of 0.27 cm2/V⋅s for transistors on glass and 0.092 cm2/V⋅s for transistors on flexible PEN. The highest transconductance of 12.2 µS was achieved for [7]phenacene transistors on glass, which was lower than that achieved for DNTT transistors. In addition, nearly hysteresis-free behaviour, improved charge carrier injection/extraction properties, and reduced threshold voltage were achieved. Finally, a semi-empirical transistor model was developed in Matlab. The model was validated using d.c. and a.c. measurements obtained on DNTT transistors with high transconductance. Four fitting parameters were extracted by optimising a fitting function using genetic algorithm. The model reproduces the d.c. transistor measurements with high accuracy. The error between the measured and simulated peak-to-peak a.c. transconductance values ranged from 1.7% to 11.6%.This thesis presents the development of low-voltage organic thin-film transistors with high transconductance. This was achieved by employing ultra-thin bi-layer gate dielectric consisting of aluminium oxide (AlOx) and a self-assembled monolayer of octadecyl phosphonic acid (C18PA) and by increasing the channel width of the transistors through the implementation of the multi-finger source/drain contacts. The transistors based on dinaphtho[2,3-b:2',3'-f]thieno[3,2-b]thiophene (DNTT) exhibited low turn-on voltage and a.c. transconductance around 30 to 60 µS. Transistor amplifiers based on such transistors exhibited voltage gain approaching 10 V/V and a gain of about 2 V/V when the supply voltage was limited to 5 V. Next, a series of [n]phenacenes ([n] = 5, 6, or 7) was used for the first time in combination with the thin AlOx/C18PA dielectric bi-layer. Regardless of the substrate and the source-drain contact geometry, the field-effect mobility of such transistors was found to increase with increasing length of the conjugated [n]phenacene core, leading to the best performance for [7]phenacene with the largest average field-effect mobility of 0.27 cm2/V⋅s for transistors on glass and 0.092 cm2/V⋅s for transistors on flexible PEN. The highest transconductance of 12.2 µS was achieved for [7]phenacene transistors on glass, which was lower than that achieved for DNTT transistors. In addition, nearly hysteresis-free behaviour, improved charge carrier injection/extraction properties, and reduced threshold voltage were achieved. Finally, a semi-empirical transistor model was developed in Matlab. The model was validated using d.c. and a.c. measurements obtained on DNTT transistors with high transconductance. Four fitting parameters were extracted by optimising a fitting function using genetic algorithm. The model reproduces the d.c. transistor measurements with high accuracy. The error between the measured and simulated peak-to-peak a.c. transconductance values ranged from 1.7% to 11.6%
Resource discovery in heterogeneous digital content environments
Get PDFThe concept of 'resource discovery' is central to our understanding of how users explore, navigate, locate and retrieve information resources. This submission for a PhD by Published Works examines a series of 11 related works which explore topics pertaining to resource discovery, each demonstrating heterogeneity in their digital discovery context. The assembled works are prefaced by nine chapters which seek to review and critically analyse the contribution of each work, as well as provide contextualization within the wider body of research literature. A series of conceptual sub-themes is used to organize and structure the works and the accompanying critical commentary. The thesis first begins by examining issues in distributed discovery contexts by studying collection level metadata (CLM), its application in 'information landscaping' techniques, and its relationship to the efficacy of federated item-level search tools. This research narrative continues but expands in the later works and commentary to consider the application of Knowledge Organization Systems (KOS), particularly within Semantic Web and machine interface contexts, with investigations of semantically aware terminology services in distributed discovery. The necessary modelling of data structures to support resource discovery - and its associated functionalities within digital libraries and repositories - is then considered within the novel context of technology-supported curriculum design repositories, where questions of human-computer interaction (HCI) are also examined. The final works studied as part of the thesis are those which investigate and evaluate the efficacy of open repositories in exposing knowledge commons to resource discovery via web search agents. Through the analysis of the collected works it is possible to identify a unifying theory of resource discovery, with the proposed concept of (meta)data alignment described and presented with a visual model. This analysis assists in the identification of a number of research topics worthy of further research; but it also highlights an incremental transition by the present author, from using research to inform the development of technologies designed to support or facilitate resource discovery, particularly at a 'meta' level, to the application of specific technologies to address resource discovery issues in a local context. Despite this variation the research narrative has remained focussed on topics surrounding resource discovery in heterogeneous digital content environments and is noted as having generated a coherent body of work. Separate chapters are used to consider the methodological approaches adopted in each work and the contribution made to research knowledge and professional practice.The concept of 'resource discovery' is central to our understanding of how users explore, navigate, locate and retrieve information resources. This submission for a PhD by Published Works examines a series of 11 related works which explore topics pertaining to resource discovery, each demonstrating heterogeneity in their digital discovery context. The assembled works are prefaced by nine chapters which seek to review and critically analyse the contribution of each work, as well as provide contextualization within the wider body of research literature. A series of conceptual sub-themes is used to organize and structure the works and the accompanying critical commentary. The thesis first begins by examining issues in distributed discovery contexts by studying collection level metadata (CLM), its application in 'information landscaping' techniques, and its relationship to the efficacy of federated item-level search tools. This research narrative continues but expands in the later works and commentary to consider the application of Knowledge Organization Systems (KOS), particularly within Semantic Web and machine interface contexts, with investigations of semantically aware terminology services in distributed discovery. The necessary modelling of data structures to support resource discovery - and its associated functionalities within digital libraries and repositories - is then considered within the novel context of technology-supported curriculum design repositories, where questions of human-computer interaction (HCI) are also examined. The final works studied as part of the thesis are those which investigate and evaluate the efficacy of open repositories in exposing knowledge commons to resource discovery via web search agents. Through the analysis of the collected works it is possible to identify a unifying theory of resource discovery, with the proposed concept of (meta)data alignment described and presented with a visual model. This analysis assists in the identification of a number of research topics worthy of further research; but it also highlights an incremental transition by the present author, from using research to inform the development of technologies designed to support or facilitate resource discovery, particularly at a 'meta' level, to the application of specific technologies to address resource discovery issues in a local context. Despite this variation the research narrative has remained focussed on topics surrounding resource discovery in heterogeneous digital content environments and is noted as having generated a coherent body of work. Separate chapters are used to consider the methodological approaches adopted in each work and the contribution made to research knowledge and professional practice
The market reaction to convertible bond issues and the determinants of bookrunner selection
Get PDFThis thesis was previously held under moratorium from 10 February 2020 until 10 February 2022.This thesis examines three major aspects of international convertible bond offerings, particularly the market reaction and its determinants, the determinants of bookrunner selection in underwriting market and the outcomes of bookrunnerselection. These issues are important for a corporate firm to design the best features of convertible bond underwriting contracts to enhance shareholders’ wealth and mitigate asymmetric information. These are relevant to the underwriters to understand whether informational advantage gained via reputation and geographic proximity could equip them if more competitive advantage in delivering better underwriting services. Furthermore, financial regulators are benefited from this study to better understand the underwriting market in convertible bond offerings. My research focuses on an overall sample size of 11,350 convertible bond offerings worldwide issued between 1984 and 2015. I analyse the market reaction for different countries, industrial classifications and stated purpose of proceeds following the announcements of convertible bond offerings. I regress the stock price reactions against firm-specific, issue-specific, market-specific, country-specific and investor protection-specific factors. I examine the determinants of domestic, regional and reputable bookrunner selection by regressing them on firm-specific, issue-specific and market-specific. I also examine whether bookrunners with economies of scale advantage are more likely to gain underwriting contracts in convertible bonds. The bookrunner performance is analysed based on the stock price reactions, underwriting fees and offering yields-at-issue. I further explore to investigate the outcome of bookrunner selection across different regions. The findings of this study have important implications for various strands of academic literature. I contribute to reveal that the stock prices of convertible bond issuers react differently to different countries, industrial classifications and stated purpose of proceeds. My results from regression analysis show that the different market reactions are significantly associated by firm-specific, issue-specific, market specific, country-specific and legal system factors. I also contribute to identify the underwriting preferences by corporate treasurers in hiring domestic, regional and reputable bookrunners. This helps underwriters to effectively gain underwriting contracts. This study contributes to two dominant literature on geographic proximate and reputable bookrunner selections in international convertible bond offerings. My findings provides important policy suggestions for issuers and investors to make a better evaluation on the outcomes of both geographical proximate and reputablebookrunners simultaneously in international convertible bond offerings.This thesis examines three major aspects of international convertible bond offerings, particularly the market reaction and its determinants, the determinants of bookrunner selection in underwriting market and the outcomes of bookrunnerselection. These issues are important for a corporate firm to design the best features of convertible bond underwriting contracts to enhance shareholders’ wealth and mitigate asymmetric information. These are relevant to the underwriters to understand whether informational advantage gained via reputation and geographic proximity could equip them if more competitive advantage in delivering better underwriting services. Furthermore, financial regulators are benefited from this study to better understand the underwriting market in convertible bond offerings. My research focuses on an overall sample size of 11,350 convertible bond offerings worldwide issued between 1984 and 2015. I analyse the market reaction for different countries, industrial classifications and stated purpose of proceeds following the announcements of convertible bond offerings. I regress the stock price reactions against firm-specific, issue-specific, market-specific, country-specific and investor protection-specific factors. I examine the determinants of domestic, regional and reputable bookrunner selection by regressing them on firm-specific, issue-specific and market-specific. I also examine whether bookrunners with economies of scale advantage are more likely to gain underwriting contracts in convertible bonds. The bookrunner performance is analysed based on the stock price reactions, underwriting fees and offering yields-at-issue. I further explore to investigate the outcome of bookrunner selection across different regions. The findings of this study have important implications for various strands of academic literature. I contribute to reveal that the stock prices of convertible bond issuers react differently to different countries, industrial classifications and stated purpose of proceeds. My results from regression analysis show that the different market reactions are significantly associated by firm-specific, issue-specific, market specific, country-specific and legal system factors. I also contribute to identify the underwriting preferences by corporate treasurers in hiring domestic, regional and reputable bookrunners. This helps underwriters to effectively gain underwriting contracts. This study contributes to two dominant literature on geographic proximate and reputable bookrunner selections in international convertible bond offerings. My findings provides important policy suggestions for issuers and investors to make a better evaluation on the outcomes of both geographical proximate and reputablebookrunners simultaneously in international convertible bond offerings
Other Modernism : underpinning the case of the history museum of Bosnia and Herzegovina
Get PDFThe thesis advances the understanding of the changing role of modern European history museums marked by engagement and outreach as modes of addressing contemporary and conflicting issues of public history. It contributes to the growing body of knowledge on the institutional uses of heritage, highlighting the case of unique and under-represented 20th-century architecture and public museum(s) of Bosnia and Herzegovina. Inspired by the communicative action concept, the research introduces a blended heritage discourse as a method to investigate the institutional role and architecture of the History Museum of Bosnia and Herzegovina, formerly known as the Museum of Revolution in Sarajevo. The transformations of the History Museum are observed through attitudes to architectural heritage and regional identity-shaping narratives, considering it as a case of embodied social energy at risk. The systematic analysis of previously inaccessible archival records on conception, construction and proposed interventions to the building, charts the field for further research, policy and practice of sustainable renovations. The research captures the key historic periods of modernisation of the urbanenvironment in Bosnia and Herzegovina, focusing on the continuity of the contextual regionalism in the work of architects sensitive to the local vernacular and the sense of place, as a unique quality within the original architectural modernism in Central Europe. Thus, it supplements the revisions of modernist discourse in the Englishspeaking academia, with an exhaustive inclusion of the sources written in Bosnian (Croatian, Serbian) languages. The research shows that the Museum in Sarajevo has an original contributionto museology and that it demonstrates remarkable adaptability and resilience, faced with societal differentiation and fragmentation. Among other, this is manifested by strategic deployment of the Museum’s status as architectural heritage, which acts as a pivotal place of resistance to the adverse impacts of systemic and governance changes, where the fragmented social narratives might be constructively reassembled.The thesis advances the understanding of the changing role of modern European history museums marked by engagement and outreach as modes of addressing contemporary and conflicting issues of public history. It contributes to the growing body of knowledge on the institutional uses of heritage, highlighting the case of unique and under-represented 20th-century architecture and public museum(s) of Bosnia and Herzegovina. Inspired by the communicative action concept, the research introduces a blended heritage discourse as a method to investigate the institutional role and architecture of the History Museum of Bosnia and Herzegovina, formerly known as the Museum of Revolution in Sarajevo. The transformations of the History Museum are observed through attitudes to architectural heritage and regional identity-shaping narratives, considering it as a case of embodied social energy at risk. The systematic analysis of previously inaccessible archival records on conception, construction and proposed interventions to the building, charts the field for further research, policy and practice of sustainable renovations. The research captures the key historic periods of modernisation of the urbanenvironment in Bosnia and Herzegovina, focusing on the continuity of the contextual regionalism in the work of architects sensitive to the local vernacular and the sense of place, as a unique quality within the original architectural modernism in Central Europe. Thus, it supplements the revisions of modernist discourse in the Englishspeaking academia, with an exhaustive inclusion of the sources written in Bosnian (Croatian, Serbian) languages. The research shows that the Museum in Sarajevo has an original contributionto museology and that it demonstrates remarkable adaptability and resilience, faced with societal differentiation and fragmentation. Among other, this is manifested by strategic deployment of the Museum’s status as architectural heritage, which acts as a pivotal place of resistance to the adverse impacts of systemic and governance changes, where the fragmented social narratives might be constructively reassembled
Mechanistic studies on three areas of organic chemistry - a combined computational and experimental approach
Get PDFThis thesis was previously held under moratorium from 01/04/2020 to 01/04/2022This Thesis contains three results chapters. The underlying aim of these project chapters is to gain a detailed understanding of reaction mechanisms and to develop new synthetic methods based on this knowledge. The three results chapters come under the following titles.Organic Super Electron Donors Made Catalytic (Chapter 2) Neutral organic super electron donors are versatile and powerful reducing agents. So far, however, it has not been possible to use such electron donors in a catalytic sense. Also, many of these donors show significant limitations when applied in radical cascade reactions. This limitation is due to the fact that these donors form relatively long-lived radical cation species upon oxidation, which interfere with the desired radical reaction. The reason for this behaviour is, that the established classes of neutral organic electron donors are (potential) double electron donors.To address these shortcomings, this Thesis reports ways to generate and utilise neutral organic single electron donors (Scheme IV-I). According to the envisaged scheme, the organic single electron donor (i) is generated as an intermediate in a radical chain reaction. The hydrogen abstraction from (iii) by the generated radical, •R’, is the chain propagation step. It becomes evident that the scheme harbours the possibility to use the single electron donor (i) catalytically. The aminal (iii) is regenerated from the benzimidazolium salt (ii) and a hydride source. Sodium borohydride was found to be the ideal hydride source.The protocol was used to perform a range of 5-exo-trig radical cyclisation reactions and its applicability is demonstrated. Based on a combined experimental and computational approach, a plausible initiation pathway based on oxygen in the air is proposed.[Graphic here - Scheme IV-I The proposed scheme for catalytic use of an organic single electron donor (i).]The KOtBu-Et3SiH Reagent System (Chapter 3) The combination of silanes and potassium tert-butoxide allows for various different transformations. Accordingly, several mechanistic pathways and key reactive species have been proposed to account for the different reactions in the literature (Figure IV-I). Hydride transfer mechanisms have been suggested to involve the silicate (iv), which is in equilibrium with potassium hydride (v) and silyl ether (vi).Plausible radical and single electron transfer mechanisms involve (vii) and (viii). Finally, hydrogen atom transfer mechanisms via (ix) cannot be fully excluded as alternatives for many of the discussed mechanisms. The formidable challenge is to probe these different domains of reactivity separately with suitable substrates. Based on a combined approach of experimental and computational investigations, strong indications are reported for single electron transfer reactivity (presumably with species (viii) acting as the single electron donor) and hydride transfer mechanisms. Additionally, plausible mechanisms for the silylation reaction of amines (x) to give the silylated species (xi) are proposed based on computational studies (Scheme IV-II).[Graphic here - Figure IV-I Proposed key species that are responsible for the reactivity of the KOtBu-Et3SiH reagent system.][Graphic here - Scheme IV-II What is the mechanism of the silylation reaction of simple amines with the KOtBu-Et3SiH (or more generally the KOtBu-Silane) reagent system?]Concerted vs. Stepwise SNAr Mechanism (Chapter 4) Over the last decades, more and more reports accumulated in the literature that suggested certain SNAr reactions follow a concerted pathway. These days, the combined impact of these investigations has reached a critical momentum and it seems appropriate to fundamentally question the long established mechanistic picture of the SNAr reaction.High level computational (wave-function based) methods are here employed to benchmark DFT functionals, based on whether these correctly predict the mechanism of a SNAr reaction to be stepwise or concerted. A reliable functional is then used to establish trends, which show what aspects (nucleophile, counter cation, leaving group, aromatic system) of the reaction at hand influence its mechanistic propensity, and in what way. Eventually this allowed estimation of when an SNAr reaction follows a concerted and when a stepwise pathway is taken.This Thesis contains three results chapters. The underlying aim of these project chapters is to gain a detailed understanding of reaction mechanisms and to develop new synthetic methods based on this knowledge. The three results chapters come under the following titles.Organic Super Electron Donors Made Catalytic (Chapter 2) Neutral organic super electron donors are versatile and powerful reducing agents. So far, however, it has not been possible to use such electron donors in a catalytic sense. Also, many of these donors show significant limitations when applied in radical cascade reactions. This limitation is due to the fact that these donors form relatively long-lived radical cation species upon oxidation, which interfere with the desired radical reaction. The reason for this behaviour is, that the established classes of neutral organic electron donors are (potential) double electron donors.To address these shortcomings, this Thesis reports ways to generate and utilise neutral organic single electron donors (Scheme IV-I). According to the envisaged scheme, the organic single electron donor (i) is generated as an intermediate in a radical chain reaction. The hydrogen abstraction from (iii) by the generated radical, •R’, is the chain propagation step. It becomes evident that the scheme harbours the possibility to use the single electron donor (i) catalytically. The aminal (iii) is regenerated from the benzimidazolium salt (ii) and a hydride source. Sodium borohydride was found to be the ideal hydride source.The protocol was used to perform a range of 5-exo-trig radical cyclisation reactions and its applicability is demonstrated. Based on a combined experimental and computational approach, a plausible initiation pathway based on oxygen in the air is proposed.[Graphic here - Scheme IV-I The proposed scheme for catalytic use of an organic single electron donor (i).]The KOtBu-Et3SiH Reagent System (Chapter 3) The combination of silanes and potassium tert-butoxide allows for various different transformations. Accordingly, several mechanistic pathways and key reactive species have been proposed to account for the different reactions in the literature (Figure IV-I). Hydride transfer mechanisms have been suggested to involve the silicate (iv), which is in equilibrium with potassium hydride (v) and silyl ether (vi).Plausible radical and single electron transfer mechanisms involve (vii) and (viii). Finally, hydrogen atom transfer mechanisms via (ix) cannot be fully excluded as alternatives for many of the discussed mechanisms. The formidable challenge is to probe these different domains of reactivity separately with suitable substrates. Based on a combined approach of experimental and computational investigations, strong indications are reported for single electron transfer reactivity (presumably with species (viii) acting as the single electron donor) and hydride transfer mechanisms. Additionally, plausible mechanisms for the silylation reaction of amines (x) to give the silylated species (xi) are proposed based on computational studies (Scheme IV-II).[Graphic here - Figure IV-I Proposed key species that are responsible for the reactivity of the KOtBu-Et3SiH reagent system.][Graphic here - Scheme IV-II What is the mechanism of the silylation reaction of simple amines with the KOtBu-Et3SiH (or more generally the KOtBu-Silane) reagent system?]Concerted vs. Stepwise SNAr Mechanism (Chapter 4) Over the last decades, more and more reports accumulated in the literature that suggested certain SNAr reactions follow a concerted pathway. These days, the combined impact of these investigations has reached a critical momentum and it seems appropriate to fundamentally question the long established mechanistic picture of the SNAr reaction.High level computational (wave-function based) methods are here employed to benchmark DFT functionals, based on whether these correctly predict the mechanism of a SNAr reaction to be stepwise or concerted. A reliable functional is then used to establish trends, which show what aspects (nucleophile, counter cation, leaving group, aromatic system) of the reaction at hand influence its mechanistic propensity, and in what way. Eventually this allowed estimation of when an SNAr reaction follows a concerted and when a stepwise pathway is taken
Android malware detection using static analysis, machine learning and deep learning
Get PDFAndroid has been a dominant mobile operating system since 2012 as shown in Figure 1. This popularity coupled with a ubiquitous usage of smartphone in all aspects of our lives, e.g. online banking, social networking, and online shopping etc. have made Android a lucrative target for malware developers.To combat the threat of malware stealing our private information, researchers have suggested various techniques for detecting Android malware. Broadly speaking, three primary techniques have been used for malware detection. Static Analysis, performed without running the application, has been used to generate signatures of malware, that can be used to differentiate between malware and benign applications. Another technique, Dynamic Analysis, has been used to create a behaviour profile of malware and benign applications by executing them in a controlled environment and monitoring their behaviour to detect malware. Hybrid Analysis has been used to utilise signatures generated from static analysis and behaviour profile created from dynamic analysis for detecting Android malware. In recent years, complementary techniques such as Machine Learning and Deep Learning have been used to extract features from the three primary analysis techniques and feed them to several algorithms for classification purposes. Deep Learning is a subfield of Machine Learning that relates to structuring algorithms in layers to mimic human neural network. The artificial neural network is used to solve complex problems using different algorithms.In this dissertation, firstly, a systematic review is presented to amalgamate current approaches for detecting Android malware, and custom-built malware detection technologies. As a result of the literature evaluation, a taxonomy is suggested for Android malware detection. Furthermore, trends in the usage of the major analytical techniques and complementary techniques are shown. Research gaps in the Android malware detection area are identified for future research direction.Secondly, Droid Fence, a custom-built web-based framework, for managing experiments is developed. Droid Fence automates the extraction of the required features from malware and benign applications directory by conducting static analysis via a frontend. Next, Droid Fence completes the automated process by storing the extracted features against each application record in a relational database, feeding them to the required machine learning and deep learning algorithms, storing the result into the database, and finally displaying the outcome of each experiment.Thirdly, developed an approach that amalgamates a set of permissions, services, and six other features (usage of https, database, dynamic code, native code, reflection, and cryptography) to generate a matrix that is used for detecting malware effectively. To the best of our knowledge, this is a novel approach that combines these features to detect malware. Droid Fence is evaluated on a dataset of 13191 applications consisting of 5787 malware and 7404 benign applications. Our results show that Droid Fence is very effective when it utilises a Sequential (Deep Learning) algorithm to detect malware, achieving accuracy, F1-measure, precision, and recall scores of 0.971, 0.967, 0.977, and 0.956 respectively. Our experiments, conducted using Droid Fence, demonstrates that deep learning Sequential algorithm scored consistently highly when compared against eight machine learning algorithms. However, the difference between the accuracy scores achieved by the Sequential (97.1%) and Random Forest Classifier (95.8%) is minimal in comparison with the remaining algorithms used in our experiments. We used a stratified k fold cross-validation method, and the result was compared for four metrics: accuracy, F1 score, precision, and recall.Finally, a conclusion and future research direction are suggested for both Android malware detection area and improvement in Droid Fence.Android has been a dominant mobile operating system since 2012 as shown in Figure 1. This popularity coupled with a ubiquitous usage of smartphone in all aspects of our lives, e.g. online banking, social networking, and online shopping etc. have made Android a lucrative target for malware developers.To combat the threat of malware stealing our private information, researchers have suggested various techniques for detecting Android malware. Broadly speaking, three primary techniques have been used for malware detection. Static Analysis, performed without running the application, has been used to generate signatures of malware, that can be used to differentiate between malware and benign applications. Another technique, Dynamic Analysis, has been used to create a behaviour profile of malware and benign applications by executing them in a controlled environment and monitoring their behaviour to detect malware. Hybrid Analysis has been used to utilise signatures generated from static analysis and behaviour profile created from dynamic analysis for detecting Android malware. In recent years, complementary techniques such as Machine Learning and Deep Learning have been used to extract features from the three primary analysis techniques and feed them to several algorithms for classification purposes. Deep Learning is a subfield of Machine Learning that relates to structuring algorithms in layers to mimic human neural network. The artificial neural network is used to solve complex problems using different algorithms.In this dissertation, firstly, a systematic review is presented to amalgamate current approaches for detecting Android malware, and custom-built malware detection technologies. As a result of the literature evaluation, a taxonomy is suggested for Android malware detection. Furthermore, trends in the usage of the major analytical techniques and complementary techniques are shown. Research gaps in the Android malware detection area are identified for future research direction.Secondly, Droid Fence, a custom-built web-based framework, for managing experiments is developed. Droid Fence automates the extraction of the required features from malware and benign applications directory by conducting static analysis via a frontend. Next, Droid Fence completes the automated process by storing the extracted features against each application record in a relational database, feeding them to the required machine learning and deep learning algorithms, storing the result into the database, and finally displaying the outcome of each experiment.Thirdly, developed an approach that amalgamates a set of permissions, services, and six other features (usage of https, database, dynamic code, native code, reflection, and cryptography) to generate a matrix that is used for detecting malware effectively. To the best of our knowledge, this is a novel approach that combines these features to detect malware. Droid Fence is evaluated on a dataset of 13191 applications consisting of 5787 malware and 7404 benign applications. Our results show that Droid Fence is very effective when it utilises a Sequential (Deep Learning) algorithm to detect malware, achieving accuracy, F1-measure, precision, and recall scores of 0.971, 0.967, 0.977, and 0.956 respectively. Our experiments, conducted using Droid Fence, demonstrates that deep learning Sequential algorithm scored consistently highly when compared against eight machine learning algorithms. However, the difference between the accuracy scores achieved by the Sequential (97.1%) and Random Forest Classifier (95.8%) is minimal in comparison with the remaining algorithms used in our experiments. We used a stratified k fold cross-validation method, and the result was compared for four metrics: accuracy, F1 score, precision, and recall.Finally, a conclusion and future research direction are suggested for both Android malware detection area and improvement in Droid Fence
The solid-state structure of tricyclic antidepressants
No full textThe active pharmaceutical ingredient (API) is the biologically active molecule present in medication. These are most commonly formulated in the crystalline solid state due to the stability and advantageous material properties generally demonstrated. The pharmaceutical industry is interested in the stability of their APIs, because it is important that it does not change during production, while on the shelf or while in the recipient’s home. Solubility is another chemical property to control for maximum efficacy of a drug, as it aids bioavailability and distribution of the API to the desired site of action. Other important properties include melting point, mechanical hardness, and particle size. These properties have a large effect on the manufacturability of APIs as they affect processes such as grinding, mixing, and tabletting of solid products. Aqueous solubility, hygroscopicity, melting point, mechanical hardness and particle size are all examples of material properties of the bulk solid of solid form APIs. They strongly depend on the number, type and strength of intermolecular interactions in the solid and are therefore properties that can be controlled and altered by changing the solid-state structure without having to alter the API molecule.There are four main ways to alter the solid-state structure and change material properties without altering the API molecule. Altering the solid-state structure affects the number, type and strength of intermolecular interactions. The options to alter these interactions include the choice between amorphous and crystalline material, investigating the polymorphic landscape of the API, and co-crystal or salt formation. Salt formation was the route that was focussed on in this work.A salt screen was carried out on four closely related tricyclic antidepressants (TCAs): amitriptyline, nortriptyline, imipramine, and doxepin. All salt formation experiments were carried out exclusively in water in order to reduce the number of variables when comparing the resulting salt forms. A dataset of 29 newly elucidated crystal structures of TCA salt forms was generated.From this dataset a selection of nortriptyline salt forms was made for an apparent aqueous solubility study. The salt forms were selected due to the counterions being ‘Generally Regarded as Safe’ (GRAS) for use in pharmaceutical industry. The salt forms used were bromide, iodide, salicylate, tosylate, besylate and oxalate. A clear trend was observed, where the salt forms with halogen counterions showed a higher solubility than organic counterions. The most soluble salt form of nortriptyline was the chloride salt which was over 200 times more soluble than the salicylate salt, which was determined to be the least soluble.Structural analysis was carried out on the dataset which found that it would appear that all four TCA bases were much more likely to form Z’ > 1 structures than is normal. Only 8.8% of structures in the Cambridge Structural Database (CSD) were found to have a Z’ value larger than 1, compared to 52% of newly elucidated TCA crystal structures from this work. Three conformational motifs were investigated as possible causes for this increased likelihood of anomalous Z’ values. Three ring/chain conformations were identified, as well as three torsion angle α, and two torsion angle β options. Indeed, in structures with higher Z’ values, API molecules with mixtures of these conformations were observed. In addition, five of the Z’ = 1 TCA salt structures contained a cation site where all atoms of the cation were disordered. There is a chance that larger super cells or modulated structures have not been identified here and that these structures could be ordered with higher Z’ values.The structures were further analysed as part of a larger data set of tricyclic antidepressants obtained from the CSD by the crystal packing similarity tool in Mercury. Four isostructural groups were identified. It appeared that the four cation packing isostructural groups could be divided into two types. Groups 1 and 2 contained different API molecules and/or very diverse counterions, and different hydration states. This illustrated that cation packing can be isostructural even where the structures that make up the groups involve different cations, different anions, and different solvation states. Group 1 seems to be a group where identical cation packing is caused by the presence of the picrate anion. In contrast groups 3 and 4 contain pairs of closely related structures which differ only in the detail of the halide anion. Here isostructural packing of the cation is simply a subset of a larger structural similarity.The crystal packing similarity analysis was repeated with a dataset of 85 carbamazepine and dihydrocarbamazepine structures found in the CSD. Here, 12 isostructural groups were identified, that could be split into three types. Type 1, groups 1 and 2, consisted of polymorphic forms. In group 1, it may be possible that the structures were not isostructural, but indeed the same form, but solved in different crystals systems. For type 2, it appeared the groups were made up of structures where the co-formers were geometrically and functionally similar to each other. Here, isostructural packing of the cation is simply a subset of a larger structural similarity, as in the second type of TCA isostructural groups. Type 3 had less in common, the structures in each group contained either dissimilar co-formers or different solvation states. The driving force of isostructural packing here seemed to be the hydrogen bonding motif and the link between this and the structures in the type 3 groups being channel solvates. Overall, in most cases in groups 3 through 12, the unit cell dimensions were very similar. This could be further studied as a tool to predict isostructural forms.The active pharmaceutical ingredient (API) is the biologically active molecule present in medication. These are most commonly formulated in the crystalline solid state due to the stability and advantageous material properties generally demonstrated. The pharmaceutical industry is interested in the stability of their APIs, because it is important that it does not change during production, while on the shelf or while in the recipient’s home. Solubility is another chemical property to control for maximum efficacy of a drug, as it aids bioavailability and distribution of the API to the desired site of action. Other important properties include melting point, mechanical hardness, and particle size. These properties have a large effect on the manufacturability of APIs as they affect processes such as grinding, mixing, and tabletting of solid products. Aqueous solubility, hygroscopicity, melting point, mechanical hardness and particle size are all examples of material properties of the bulk solid of solid form APIs. They strongly depend on the number, type and strength of intermolecular interactions in the solid and are therefore properties that can be controlled and altered by changing the solid-state structure without having to alter the API molecule.There are four main ways to alter the solid-state structure and change material properties without altering the API molecule. Altering the solid-state structure affects the number, type and strength of intermolecular interactions. The options to alter these interactions include the choice between amorphous and crystalline material, investigating the polymorphic landscape of the API, and co-crystal or salt formation. Salt formation was the route that was focussed on in this work.A salt screen was carried out on four closely related tricyclic antidepressants (TCAs): amitriptyline, nortriptyline, imipramine, and doxepin. All salt formation experiments were carried out exclusively in water in order to reduce the number of variables when comparing the resulting salt forms. A dataset of 29 newly elucidated crystal structures of TCA salt forms was generated.From this dataset a selection of nortriptyline salt forms was made for an apparent aqueous solubility study. The salt forms were selected due to the counterions being ‘Generally Regarded as Safe’ (GRAS) for use in pharmaceutical industry. The salt forms used were bromide, iodide, salicylate, tosylate, besylate and oxalate. A clear trend was observed, where the salt forms with halogen counterions showed a higher solubility than organic counterions. The most soluble salt form of nortriptyline was the chloride salt which was over 200 times more soluble than the salicylate salt, which was determined to be the least soluble.Structural analysis was carried out on the dataset which found that it would appear that all four TCA bases were much more likely to form Z’ > 1 structures than is normal. Only 8.8% of structures in the Cambridge Structural Database (CSD) were found to have a Z’ value larger than 1, compared to 52% of newly elucidated TCA crystal structures from this work. Three conformational motifs were investigated as possible causes for this increased likelihood of anomalous Z’ values. Three ring/chain conformations were identified, as well as three torsion angle α, and two torsion angle β options. Indeed, in structures with higher Z’ values, API molecules with mixtures of these conformations were observed. In addition, five of the Z’ = 1 TCA salt structures contained a cation site where all atoms of the cation were disordered. There is a chance that larger super cells or modulated structures have not been identified here and that these structures could be ordered with higher Z’ values.The structures were further analysed as part of a larger data set of tricyclic antidepressants obtained from the CSD by the crystal packing similarity tool in Mercury. Four isostructural groups were identified. It appeared that the four cation packing isostructural groups could be divided into two types. Groups 1 and 2 contained different API molecules and/or very diverse counterions, and different hydration states. This illustrated that cation packing can be isostructural even where the structures that make up the groups involve different cations, different anions, and different solvation states. Group 1 seems to be a group where identical cation packing is caused by the presence of the picrate anion. In contrast groups 3 and 4 contain pairs of closely related structures which differ only in the detail of the halide anion. Here isostructural packing of the cation is simply a subset of a larger structural similarity.The crystal packing similarity analysis was repeated with a dataset of 85 carbamazepine and dihydrocarbamazepine structures found in the CSD. Here, 12 isostructural groups were identified, that could be split into three types. Type 1, groups 1 and 2, consisted of polymorphic forms. In group 1, it may be possible that the structures were not isostructural, but indeed the same form, but solved in different crystals systems. For type 2, it appeared the groups were made up of structures where the co-formers were geometrically and functionally similar to each other. Here, isostructural packing of the cation is simply a subset of a larger structural similarity, as in the second type of TCA isostructural groups. Type 3 had less in common, the structures in each group contained either dissimilar co-formers or different solvation states. The driving force of isostructural packing here seemed to be the hydrogen bonding motif and the link between this and the structures in the type 3 groups being channel solvates. Overall, in most cases in groups 3 through 12, the unit cell dimensions were very similar. This could be further studied as a tool to predict isostructural forms
Atmospheric re-entry analysis of space objects
No full textThis Ph.D. work focuses on the investigation of numerical methods for estimating the aerodynamics and aerothermodynamics of space objects re-entering the Earth atmosphere. The well-known simplified panel methods applicable to hypersonic aerothermodynamics within continuum and rarefied flow regimes have been implemented in a numerical framework in order to assess different methods and techniques aiming at increasing their accuracy on arbitrarily shaped objects and assemblies. The main tested methods and techniques comprise: the use of a combination of panels visibility detection algorithms, the application of a local radius computation and smoothing algorithm, the introduction of high fidelity-based correction factors, and the definition of general rarefied transitional regime bridging functions for both aerodynamic and aerothermodynamics. Various re-entry scenarios, experimental tests, and literature references have been simulated with the proposed methods in order to understand their accuracy. A preliminary validation of the aerodynamics and aerothermodynamics was performed on different test cases: Space Shuttle Orbiter, Orion Crew Exploration Vehicle, Mars Microprobe, Stardust Sample Return Capsule, VEGA payload fairing, and various other test cases such as spheres, and blunted cones. A particular effort was made to define and apply the methods for allowing the simulation of complete atmospheric re-entry scenarios, in order to simulate and study controlled and uncontrolled re-entres of simple and complex objects assemblies.;In order to evaluate the applicability of the proposed methodologies, different re-entry scenarios have been simulated, such as: Stardust SRC (focusing on the thermal protection system ablation and temperature), and the Intermediate Experimental Vehicle (estimating the wall temperature over the re-entry). The last application which has been studied was the estimation of the re-entry human casualty risk as expressed by the space debris mitigation guidelines. For such application, a toy satellite was built and simulated using a statistical approach based on the use of surrogate models and Monte Carlo technique aiming at providing a statistical impact ground footprint along with the human casualty risk distribution for a set of uncertain parameters given as input.This Ph.D. work focuses on the investigation of numerical methods for estimating the aerodynamics and aerothermodynamics of space objects re-entering the Earth atmosphere. The well-known simplified panel methods applicable to hypersonic aerothermodynamics within continuum and rarefied flow regimes have been implemented in a numerical framework in order to assess different methods and techniques aiming at increasing their accuracy on arbitrarily shaped objects and assemblies. The main tested methods and techniques comprise: the use of a combination of panels visibility detection algorithms, the application of a local radius computation and smoothing algorithm, the introduction of high fidelity-based correction factors, and the definition of general rarefied transitional regime bridging functions for both aerodynamic and aerothermodynamics. Various re-entry scenarios, experimental tests, and literature references have been simulated with the proposed methods in order to understand their accuracy. A preliminary validation of the aerodynamics and aerothermodynamics was performed on different test cases: Space Shuttle Orbiter, Orion Crew Exploration Vehicle, Mars Microprobe, Stardust Sample Return Capsule, VEGA payload fairing, and various other test cases such as spheres, and blunted cones. A particular effort was made to define and apply the methods for allowing the simulation of complete atmospheric re-entry scenarios, in order to simulate and study controlled and uncontrolled re-entres of simple and complex objects assemblies.;In order to evaluate the applicability of the proposed methodologies, different re-entry scenarios have been simulated, such as: Stardust SRC (focusing on the thermal protection system ablation and temperature), and the Intermediate Experimental Vehicle (estimating the wall temperature over the re-entry). The last application which has been studied was the estimation of the re-entry human casualty risk as expressed by the space debris mitigation guidelines. For such application, a toy satellite was built and simulated using a statistical approach based on the use of surrogate models and Monte Carlo technique aiming at providing a statistical impact ground footprint along with the human casualty risk distribution for a set of uncertain parameters given as input
Access Repository Dashboard
Do you manage STAX (Strathclyde Repository)? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!