School of Aerospace, Transport and Manufacturing (SATM)
Permanent URI for this community
Browse
Browsing School of Aerospace, Transport and Manufacturing (SATM) by Course name "Manufacturing"
Now showing 1 - 17 of 17
Results Per Page
Sort Options
Item Open Access Advanced flow technologies for the controlled & continuous manufacture of nanoscale materials(Cranfield University, 2019) Isaev, Svetlin; Makatsoris, Charalampos (Harris)Batch processes have been successfully used in the process industry over two centuries. However, changing customer demands and discovery of novel products have led the scientists and engineers to develop new manufacturing methods for the process industry. High-value products such as nanomaterials, smart and functional materials require precise process control for the entire product. Controlling of particle size and shape becomes more difficult in the large scale batch processes. Therefore, over the past few decades, there has been an increasing interest in the flow processing techniques due to their inherent benefits, such as better heat and mass transfer and small control volumes. Continuous Oscillatory Baffled Reactor (COBR) is a novel type of flow reactor. COBR combines oscillatory motion and periodically placed baffled flow channels to generate plug flow conditions, providing better mixing control similar to microreactors. Plug flow conditions can be achieved with the combination of optimum net flow, oscillatory amplitude and frequency using COBRs. With this new reactor and mixing concept, high-value products can be manufactured more efficiently using uniform mixing conditions and better temperature control. This will decrease the reaction time and production cost of novel products, use less energy, and increase time-to-market of novel products. The aim of this research is to develop a scalable and continuous manufacturing platform using continuous oscillatory baffled reactors to produce high-value products in low cost. The focus of this study includes developing modular oscillatory baffled reactors, characterisation of modular oscillatory baffled reactors using experimental methods, developing scale-up methodology from laboratory scale to industrial production size and demonstration of nanomaterial synthesis using modular oscillatory flow reactor...[cont.]Item Open Access Analysis of the evolution of aerospace manufacturing ecosystems(Cranfield University, 2020-06) Luna Andrade, Jose Junior; Salonitis, Konstantinos; Brintrup, AlexandraThe aerospace manufacturing industry is predicted to continue growing. Understanding its evolution is thus essential to prepare optimal conditions to nurture its growth. This research aims to help the growth of emerging aerospace ecosystems by identifying evolution patterns and categorising key enablers that have encouraged the growth of developed ones. The term aerospace ecosystem is used to embrace all the business activities and infrastructure that are related to the entire aerospace’s supply chain in a specific country. Inspired by studies that have successfully combined economics and network science, in this research, bipartite country-product networks are developed based on trade data over 25 years. The United Kingdom (UK), the United States of America, France, Germany, Canada and Brazil’s are first analysed as evidence suggests that their aerospace ecosystems are within the most developed in the world. Then, China and Mexico’s networks are analysed and compared with developed ones, as these countries have evidenced emergent aerospace ecosystems. Results reveal that developed ecosystems tend to become more analogous, as countries lean towards having a revealed comparative advantage (RCA) in the same group of products. Further analysis shows that manufactured products have a stronger correlation to an aerospace ecosystem than primary products; and in particular, the automotive sector shows the highest correlation with positive aerospace sector evolution. Key enablers related to the growth of the UK and Mexico’s aerospace ecosystems are identified and categorised using interpretive structural modelling (ISM) and cross-impact matrix multiplication applied to classification (MICMAC) methodologies. Results evidence relevant differences in the categorisation of key enablers among a developed and emergent aerospace ecosystems. On the other hand, it was identified that geopolitical factors and the automotive ecosystem are underpinning enablers for both aerospace ecosystem’s evolution. The final aim is that results of this research could be implemented on emerging aerospace ecosystems by emulating the patterns and key enablers that have characterised the evolution of developed aerospace ecosystems.Item Open Access Design of redistributed manufacturing networks: a model-based framework(Cranfield University, 2020-10) Haddad, Yousef; Salonitis, Konstantinos; Emmanouilidis, ChristosThroughout the last century, manufacturing has been characterised by mass production conducted in central facilities, benefitting from economies-of-scale. These central facilities supply, and are supplied by sprawling, complex supply chains that are slow to adapt to demand changes and supply disruptions. As production strategies are gradually shifting from economies-of-scale to economies-of-scope to cater for increasingly complex heterogeneous demand and shorter product life cycles, new configurations are required to enable manufacturing systems to accommodate these demand changes efficiently. One area that has the potential to improve the responsiveness of manufacturing systems is redistributed manufacturing (RdM). RdM is a manufacturing paradigm where production is performed in a network of small, autonomous and geographically distributed facilities. Motivated by the potential opportunities that RdM could bring, this thesis develops a model-based decision-making framework for the design and operation of RdM networks. The framework is context-independent, addresses strategic, tactical and operational decision-making levels and accounts for the interdependence between these decisions in a stochastic environment. The framework is validated methodically through computational experiments on two case studies of different natures and objectives. Experts opinions were solicited throughout the design stage of this research, the implementation of the case studies and the analysis of the results. Results reveal that even when the objectives of the modelled systems are substantially different, the framework generates consistent outputs. The main takeout from the experiments’ results is that the RdM paradigm consistently produces significantly better service level performance, demonstrated by fewer occurrences of unmet demands and shorter lead times. However, although sufficiently close, the RdM paradigm is not as cost efficient as the traditional centralised manufacturing paradigm.Item Open Access Development of a context-aware internet of things framework for remote monitoring services(Cranfield University, 2020-12) Al-Shdifat, Ali M. A.; Emmanouilidis, Christos; Starr, AndrewAsset management is concerned with the management practices necessary to maximise the value delivered by physical engineering assets. Internet of Things (IoT)-generated data are increasingly considered as an asset and the data asset value needs to be maximised too. However, asset-generated data in practice are often collected in non-actionable form. Moreover, IoT data create challenges for data management and processing. One way to handle challenges is to introduce context information management, wherein data and service delivery are determined through resolving the context of a service or data request. This research was aimed at developing a context awareness framework and implementing it in an architecture integrating IoT with cloud computing for industrial monitoring services. The overall aim was achieved through a methodological investigation consisting of four phases: establish the research baseline, define experimentation materials and methods, framework design and development, as well as case study validation and expert judgment. The framework comprises three layers: the edge, context information management, and application. Moreover, a maintenance context ontology for the framework has developed focused on modelling failure analysis of mechanical components, so as to drive monitoring services adaptation. The developed context-awareness architecture is expressed business, usage, functional and implementation viewpoints to frame concerns of relevant stakeholders. The developed framework was validated through a case study and expert judgement that provided supporting evidence for its validity and applicability in industrial contexts. The outcomes of the work can be used in other industrially-relevant application scenarios to drive maintenance service adaptation. Context adaptive services can help manufacturing companies in better managing the value of their assets, while ensuring that they continue to function properly over their lifecycle.Item Open Access The development of a design and development framework between OEM and supplier(Cranfield University, 2019-07) Ogundana, Damilola; Al-Ashaab, Ahmed; McLaughlin, PatrickSuppliers are of great importance to OEMs because of the benefits that can be gained from collaborating with them. But for the OEM to select the appropriate supplier for the specific job they want accomplish, they must first create criteria that can be used to evaluate the supplier. The purpose of this thesis is to develop a design and development framework between OEM and supplier. The framework is focused only on the design and development activities and not manufacturing. This research was able to identify the right criteria for OEMs to use to assess, select and evaluate suppliers. Moreover, it was able to clarify the difference in criteria for each of the three aforementioned activities. The construction of the framework commenced with the use of an extensive literature review which was followed by an industrial field study consisting of 5 interviews with four companies who specialise in different sectors of engineering. The outcomes were integrated to generate the contents of the supply chain framework. A case study was simulated in order to verify the framework. The design and development framework provides the necessary means by which an OEM can assess, select and evaluate suppliers during product design and development processes. As a result of this, a functionally feasible and enhanced design and is more efficient can be realised. The framework that was developed as a result of this is very comprehensive and is able to mitigate the challenges faced in the industry today, regarding a outsourcing of OEMs’ product development activities to supplier. The contributions to the knowledge are as follows: (1) The developed framework provides a clear understanding of what constitutes as assessment criteria, selection criteria and evaluation criteria in product design and development within the supply chain. (2) The framework mitigates the evolving challenges faced by OEM and suppliers when product design and development is outsourced. (3) The developed framework encompasses all the activities involved in assessing, selecting and evaluating suppliers throughout the outsourcing process.Item Open Access Employee performance modelling using system dynamics(Cranfield University, 2019-09) Alefari, Mudhafar Rashed Saeed All Warea; Salonitis, KonstantinosEmployee performance is something dynamic, but can have great impact on the overall performance of any company. This is understood by companies and human resource management departments are responsible for measuring the performance of the employees, and come up with ideas on how to improve this constantly. Such practices include training of employees, providing initiatives such as bonuses and day offs. Furthermore, the literature review has highlighted that leadership style can have a great impact as well. Looking in the literature of employee performance, it was clear as well that there has not been a model that can be used for predicting the impact of such initiatives from systems point of view. So the initial aim was to develop such a model that can help manufacturing companies better handle the dynamic nature of employee performance and if possible help with the decision making when deciding which initiatives to be introduced. The literature review was focused in identifying the factors that have an impact on the employee performance and their possible interrelations. Then the best modelling approach was investigated. Modelling techniques such as discrete event simulation, agent based modelling and system dynamics were considered, with the latter selected at the end as the focus is on the impact of the change of policies and not the individual employees who cannot be modelled due to the random way of their behaviour. System dynamics models were developed based on this analysis and collecting data protocols were formulated for collecting information from companies. The models were validated in two companies in UAE. They can predict the impact that specific changes in policies will have in the employee performance and can guide the companies about what changes they should introduceItem Open Access A framework for the development of learning management systems for higher education institutions in the Kingdom of Saudi Arabia(Cranfield University, 2020-08) Alduraywish, Yousef Ahmed Y; Salonitis, Konstantinos; Patsavellas, JohnThis study focuses on a framework for the development of the Learning Management System (LMS) in the Kingdom of Saudi Arabia (KSA) higher education institutes (HEIs) from information systems (IS) perspective, using the Technology Acceptance Model (TAM) and Design Reality Gap (ITPOSMO) model. The research methodology consists of six stages which adopts the paradigm of pragmatism and the research design of mixed-methods. A case study design is used to investigate the implementation of LMS in the Al-Imam Mohamed bin Saud Islamic University. The quantitative part was designed to investigate the attitude of users towards the usefulness of the LMS and to assess the acceptance level of LMS among university users. The qualitative part was designed to explore the gap between the proposed implementation of the LMS and reality. The survey received valid responses from 129 academic and 1548 student. A semistructured interviews with 21 participants. The sample was achieved via a purposive sampling technique. The quantitative data was analysed using descriptive statistical analysis and correlation coefficient. The qualitative data were analysed using a thematic analysis approach. The study identifies the barriers influencing effective LMS in KSA HEIs as 1) technology barriers (lack of IT infrastructure, incomplete functionalities and lack of integration); 2) human barriers (lack of knowledge of the importance of elearning, lack of expertise and competencies); and 3) organisation barriers (organisational preparedness, unclear of requirements, lack of training, resistance and financial constraints). The contribution of this research includes a new model derived from the ITPOSMO model and TAM to investigate LMS in the context of real circumstances, and the physical environment that exists in KSA HEIs. The research focus is more on meso level while encompassing first and third levels as reference for better understanding (Richter et al., 2009). The results lead to developing a framework for the development of LMS in KSA HEIs.Item Open Access Framework to assess the maturity level of learning analytics in higher education and drive learning services improvement(Cranfield University, 2020-05) Alenezi, Abdullah; Emmanouilidis, Christos; Al-Ashaab, AhmedThis research was aimed at developing a framework that could be utilised to assess the maturity level of learning analytics (LA) in virtual learning environment (VLE) in higher education institutions (HEI). The assessment of the maturity level of LA in VLE in HEI contributes to enhancing the educational learning programmes and academic services offering to the learners. The successful implementation of LA in an HEI could help improve teaching and learning processes, thereby improving students’ learning experiences (Larrabee Sønderlund et al., 2019; Sclater et al., 2016; Waheed et al., 2020). However, most HEIs often do not know where to start from in implementing programmes for using VLE and LA; thus, the contribution of this study to offer guidance for HEIs. In order to develop the LA maturity assessment framework, a multi-phases methodological approach was adopted which involved 6 key phases (understanding the literature, a field study to gain a high-level perspective of VLE and LA, development of LA maturity model, development of a performance measurement tool, formulation of road map recommendations, and case study validation and expert judgment). The developed LA maturity model comprises of five levels: basic (level 1), developing (level 2), functional (level 3), advanced (level 4) and optimised (level 5). In determining these LA maturity level, the performance measurement tool has to be applied. This performance measurement tool assesses an HEI’s performance in four key components of LA: process, infrastructure, data and human resources and skills. The LA maturity model and performance measurement tool facilitate the road map recommendations. Based on an HEI’s assessed LA maturity level, recommendations are suggested on how progress can be made in LA implementation. The developed LA assessment framework was validated through case studies (PAAET and Cranfield University) and expert judgement that proved its validity and application to different educational contexts. The case study validation showed the differences in performance scores and maturity levels of the two HEIs with specific recommendations relevant to each context being made. Expert judgements highlighted the contribution of the framework to LA which is a relatively new area of research.Item Open Access Fundamental investigation understanding casting of lead sheet(Cranfield University, 2020-04) Prabhakar, Arun; Jolly, Mark R.; Salonitis, KonstantinosLead sheet is widely used in the construction industry for roofing and flashing applications. The roots of this process can be tracked back to the Roman times when sandcast lead sheets were used for a wide variety of applications. Sandcast lead sheets are characterised by their superior aesthetic performance and mottled appearance. These days such sheets are used for premium roofing and flashing applications in the heritage construction industry. Lead sheet is also manufactured using a type of continuous casting process also called as the ‘Direct Method (DM)’. This thesis focuses on a fundamental investigation of both these processes used for manufacture of cast lead. Just like any casting process, sand casting of lead sheet suffers from the presence of surface defects. In this study, a surface defect type, hereby referred to as ‘grooves’, has been investigated. The focus has been laid on the identification of the main factors affecting defect formation in this process. Based on a set of screening experiments performed using Scanning Electron Microscopy (SEM) as well as the existing literature, a number of factors affecting the formation of such defects was identified and their corresponding significance was estimated. Two-dimensional Computational Fluid Dynamics (CFD) simulations have been performed to simulate the melt flow and solidification stages of the lead sandcasting process. The effects of process parameters such as pouring temperature, screed velocity and clearance between the screed and the sandbed on the final quality of the lead sheet are investigated. Sheet quality is quantified by measuring the variance and the average of the final sheet thickness over the sandbed length. The CFD model has been validated against experimental results by comparing the evolution of the lead-sandbed interface temperature against data collected by thermocouples during the evolution of the process. The direct method of casting lead is a much more energy efficient compared to the conventional rolling process which requires a casting process before rolling to achieve the required thickness. This work also looks into the energy consumption in different stages of the DM process and suggests pointers for improvement. An energy audit of the process is conducted, and the consumption is analysed at different stages and compared with rolled lead. A two-dimensional numerical model of the DM process was developed and different process parameters affecting the thickness of the final cast sheet is studied. Effects of parameters like volume flow rate, heat transfer coefficient, speed of rotation of the casting drum and its immersion are investigated. The studies were conducted in collaboration with ML Operations, a cast lead sheet manufacturer based in Derbyshire and the findings of the study were implemented successfully.Item Open Access Investigation of productivity, energy efficiency, quality and cost for laser drilling(Cranfield University, 2020-10) Sarfraz, Shoaib; Shehab, Essam; Salonitis, Konstantinos; Suder, WojciechLaser drilling is a high speed, non-contact advanced machining process and has proven to be an important industrial process for producing cooling holes in various aeroengine components; in particular high-pressure turbine blades, combustor liners and nozzle guide vanes. However, an increase in the number of cooling holes demands the need for effective utilisation of laser drilling process capability. Material removal rate (MRR), specific energy consumption (SEC), hole taper and the drilling cost are the basic performance indicators to meet this goal. Hence, this research aims to examine the laser drilling process in terms of the mentioned performance measures. Taking into account the significance of material removal quantity, energy efficiency, product quality and manufacturing cost, this study is performed in the form of an experimental investigation for three laser drilling processes, namely, single-pulse drilling, percussion and trepanning. Two different laser drilling setups were prepared to produce holes in Inconel 718 superalloy sheets using flashlamp-pumped Nd:YAG laser and Quasi-CW fibre laser. This research contributes to an evaluation of the influence of laser drilling process parameters on the MRR, SEC, hole quality and drilling cost. Moreover, the performance of laser drilling methods has been compared in relation to the selected performance measures. To further understand the significance of laser sources, the performance of laser drilling was compared for the mentioned drilling setups. This research also introduced a detailed cost analysis to explore the economic implications of the laser drilling process. In addition, optimal drilling conditions were determined aiming to maximise the MRR and minimise hole taper and drilling cost.Item Open Access Ontology-based augmented reality content-related techniques and their impact in knowledge capture and re-use within maintenance diagnosis(Cranfield University, 2020-05) Fernandez Del Amo Blanco, Inigo; Erkoyuncu, John AhmetThis PhD thesis aims to study ontology-based AR content-related methods and their impact in knowledge transfer, capture and re-use for cost-effective human knowledge integration in digital diagnostic systems. Industry 4.0 has revealed the importance of maintainers’ knowledge capture and re-use in diagnostics systems for providing satisfactory solutions in cases where those systems cannot (e.g. nofault-found). Augmented Reality (AR) utilises content-related techniques to transfer knowledge to maintainers for improving efficiency and effectiveness of diagnosis tasks. Academic literature has shown that AR can also be utilised for knowledge capture and re-use, but this has only been demonstrated in simple, step-by-step repair operations. In diagnosis research, ontology-based methods are applied to capture and re-use knowledge from unstructured and heterogenous sources like humans. Nevertheless, these methods have not made use of AR potential to contextualise knowledge and so, improve efficiency and effectiveness of knowledge capture and re-use diagnosis operations...[cont.]Item Open Access Physics-based modelling of cyclic deformation and microstructure-sensitive fatigue crack propagation from shallow scribes(Cranfield University, 2020-12) Ashraf, Farhan; Castelluccio, Gustavo M.; Khan, Muhammad AliFace-centered cubic (FCC) metals with low to medium stacking fault energy (SFE) develop similar mesoscale substructures under cyclic loading. The formation of these substructures is controlled by dislocation interactions and loading conditions. For instance, cross slip facilitates cell formation and Hirth locks define the labyrinth structure. In the case of aluminium (high SFE metal), cross slip is easily activated and a cell structure is often observed. However, it is not always recognised that aluminium can also form PSBs at low temperatures. This highlights that the underlying mechanism controlling the cyclic response in aluminium is not different from other FCC metals. This work proposes the role of mesoscale substructure as a material-invariant among FCC metals to predict the cyclic response of aluminium. The effect of number of cycles on modelling dislocation substructures is explored, which is found to trigger a change in dislocation structures in aluminium at 298K. A crystal plasticity framework based on mesoscale substructures is developed to study the cyclic response of aluminium under different crystal orientations, strain amplitudes, number of cycles, and temperatures. Finally, this work implemented the crystal plasticity model to study the microstructure-sensitive crack propagation from shallow scribes in pure aluminium. The gradient of fatigue indicator parameters (FIPs) is estimated as crack extends inside a grain with explicit microstructure simulations, which followed the same decaying trend predicted by experiments. Thereby, an engineering solution is proposed to couple microstructural and geometric gradients at the crack tip independently. The model predicted the transgranular fatigue life with independently coupled gradients that agree well with experiments.Item Open Access Revitalization of the embroidery industry using advanced technology in Saudi Arabia(Cranfield University, 2020-02) Algamdy, Hind Mosfeer; Khan, Muhammad Ali; Aria, IndratThe purpose of this study was to revitalize the traditional embroidery industry of the Hejaz region of the Kingdom of Saudi Arabia (KSA) by evaluating the possibility of using advanced technology, such as three-dimensional printing (3DP), in its manufacturing process. A mixed method methodology underpins this research in terms of collecting, processing and testing the data. An initial literature review revealed that factors such as an inability to meet current fashion trends, threats from global brands, lack of support from government and insufficient consumer interest are key challenges facing the traditional embroidery industry. Further qualitative evaluation pinpointed a lack of development in the manufacturing techniques of traditional embroidered clothing in the KSA as a key threat. An evaluation of existing technologies revealed that embroidery sewing machines attached to computer-aided design (CAD) software is the technology currently in use in the industry. However, due to inflexibility in adjusting to the demands of customization, it has not been able to mark any significant change. To this end, the current study proposed the use of the 3DP technique in the manufacturing process of traditional embroidered clothing but found that although 3DP has been used in the fashion industry, its use has been questioned because of wearability and quality concerns. An evaluation of responses collected from 16 manufacturers attending the Souq Okaz Festival in the city of Taif, along with 45 responses from customers of traditional embroidered clothing in three different universities in the KSA, found that both sets of stakeholders showed concerns regarding the wearability of 3DP garments. The manufacturers also shared their concerns regarding the ease of use of the technology. As a backdrop to these findings, two experiments were conducted: a washing test and a peel tensile test. The washing test revealed that 3DP embroidery designed and printed on silk, cotton and organza showed no impact from washing upon their brightness, roughness, shape or edge. However, the peel test revealed that, due to its irregular texture and shape, organza showed inconsistent adhesion of 3DP material comparative to cotton and silk. This led to the conclusion that 3DP embroidery objects present good long-term wearability, as long as the printing parameters are set up to meet the fabric architecture. Suitability, acceptability and feasibility in relation to the financial, human resource (ease of use) and supply chain aspects of 3DP embroidery clothing were also substantiated.Item Open Access Scalability of a robotic inspection and repair system(Cranfield University, 2019-12) Ortega Jarrin, Washington Cristobal; Starr, Andrew; Durazo, IsidroShift2Rail and In2Smart are two initiatives that will be part of the development of the necessary technologies to complete the Single European Railway Area (SERA). The target of this proposal is to accelerate the integration of new and advanced technologies into innovative rail product solutions. Shift2Rail has a robust framework to meet ambitious objectives. The most important is to double the capacity of the European rail system and increase its reliability and service quality by 50% while having life-cycle costs. In2Smart, as a project directed mainly of Network Rail, is measured in Technology Readiness Levels (TRL). These levels will indicate the maturity of technology for the application into the industry. The intention of this project is to reach a homogeneous TRL 3/4 demonstrator of a system capable to secure proper maintenance of rails, which is a Robotic Inspection and Repair System (RIRS). This research is focused on the scalability of the RIRS, taking into consideration the creation of a representative demonstrator that will authenticate the concept, the validation and verification of that demonstrator and finally the simulation of a scale-up system that will be more robust and will upgrade the TRL. This document contains the development of the control diagrams and schematics for the future incorporation of this control to a higher TRL prototype. The initial demonstrator consists of an autonomous railway vehicle equipped with a robotic arm that will scan the rails searching for faults and simulate a repairing process with a 3D printed polymer. The V&V of the physical demonstrator was a result of tests in the laboratory and the display of the demonstrator in several conferences and events.Item Open Access A systematic design approach to IOT security for legacy production machinery(Cranfield University, 2020-03) Tedeschi, Stefano; Emmanouilidis, Christos; Salonitis, Konstantinos; Mehnen, JornThe Internet of Things (IoT) is an emerging topic of rapidly growing technical importance for the industry. The aim is to connect objects with unique identifiers and combine them with internet connectivity for data transfer. This advanced connectivity has significant potential in the workshop-level upgrade of existing legacy equipment to unlock new features and economic benefits especially for monitoring and control applications However, the introduction of the Industrial Internet of Things (IIoT) brings new additional security and integrity risks for the industrial environment in the form of network, communication, software and hardware security risks. This thesis addresses such fundamental new risks at their root by introducing a novel approach for IoT-enabled monitoring of legacy production machinery, which consist of five stages, incorporating security by design features. The first two phases of this novel approach aim to analyse current monitoring practices and security and vulnerability issues related to the application domain. The proposed approach applies three more stages which make the domain-relevant analysis to become application specific. These include a detailed model of the application context on legacy production machinery monitoring, together with its interfaces and functionality, implementing threat mitigations combined with a new modular IoT DAQ unit mechanism, validated by functional tests against Denial of Service (DoS) and clone attacks. Thus, to be effective, the design approach is further developed with application-specific functionality. This research demonstrates an instance of this innovative riskaverse design thinking through introducing an IoT device design which is applicable to a wide set of industrial scenarios. A practical showcase example of a specific implementation of the generic IoT design is given through a concrete industrial application that upgrades existing legacy machine tool equipment. The reported work establishes a novel viewpoint for the understanding of IoT security risks and their consequent mitigation, opening a new space of riskaverse designs that can bring significant confidence in data, safety, and security of IoT-enabled industry.Item Open Access Three-dimensional subsurface defect reconstruction for industrial components using pulsed thermography(Cranfield University, 2020-05) Sirikham, Adisorn; Zhao, Yifan; Mehnen, YornPulsed thermography is a promising method for detecting subsurface defects, but most pulsed thermographic inspection results are represented in the form of 2D images. Such a representation can limit the understanding of where the defects initiate and how they grow by time, which is a key to predict the remaining use of life of component and feedback to the design to avoid such defects. Threedimensional subsurface defect visualisation is a solution that can unlock this limitation. A straightforward approach to reconstruct 3D subsurface defect is conducting two inspections on both front and rear sides. However, the deployment of this approach can be limited because 1) one side of the inspected component could be inaccessible; 2) the accuracy of measurement could be compromised if the defect thickness is very thin due to extreme closed values of defect depths from two inspections; and 3) if the defect is too deep for one side, the defect could be missed. Addressing the challenge of 3D subsurface defect reconstruction and visualisation, this thesis proposes a novel technique to measure defect depth and estimate defect thickness simultaneously through estimating the thermal wave reflection coefficient value achieved by introducing a modified heat transfer model based on a single-side inspection method. The proposed method is validated through model simulations, experimental studies, and a use case. Four composite samples with different defect types, sizes, depths and thicknesses, are used for experimental studies; a steel sample with a ‘s’ shape triangular air-gap inside is used for a use case. The simulation results show that under the noise level of 25 dB, the percentage error of the developed depth measurement method is 0.25% whilst the minimum error of the best existing method is 2.25%. From the experimental study results, the averaged percentage error of the defect thickness estimation is less than 10% if the defect thickness is no more than 3 mm. For the use case, the reconstructed defect shape is similar to the X-ray image.Item Open Access Web robot detection using supervised learning algorithms(Cranfield University, 2020-06) Chen, Hanlin; He, Hongmei; Starr, AndrewWeb robots or Web crawlers have become the main source of Web traffic. Although some bots perform well, such as search engines, other bots can perform DDoS attacks, posing a huge threat to websites. The project aims to develop an offline system that can effectively detect malicious web robots, which is not only conducive to network traffic cleaning, but also conducive to improving the network security of IoT systems and services. A comprehensive literature review for the years 2010-2019 was conducted to identify the research gap. The key contributions of the research are: 1) it provided a systematic methodology to address the web robot detection problem based on the log file from industrial company; 2) it provided an approach of feature engineering, thus overcoming the challenge of curse of dimensionality; 3) It made a big progress in the accuracy of off-line web robot detection through a holistic study on the three types of machine learning techniques based on real data from industry. Three algorithms based on Keras sequential model, random forest, and SVM, were developed with python to detect web robots from human visitors on the TensorFlow 2.0 platform. Experimental results suggested that random forest obtained the best performance in accuracy and training time...[cont.]