Dissertations / Theses on the topic 'Damage modelling'

The work presented in this thesis is a detailed study of impact damage subjected to tensile and compressive loading to determine the stiffness reduction of the damage region and identify the damage mechanisms and important parameters that control the severity of the stiffness reduction. Once the important damage parameters had been indentified a homogenised non-linear soft inclusion model was developed. This represents the mechanical and material behaviour of an impact damage region under tensile and compressive loading in a simple and easy to implement user material format. The influence of different damage parameters was determined by building ply level models of idealised impact damage with delaminations at every ply interface and fibre fracture cracks within the plies. Parametric studies were conducted on the shape and size of delaminations and on crack density and crack distribution under tensile and compressive loading and these models were validated against experimental results. In order to determine the influence of fractured fibres on the residual compressive stiffness the fibres were modelled at the micro scale with individual fibres embedded in an elastic-plastic matrix.

In this thesis, the mechanical behavior of a timber joint has been studied. The main task is to model the mechanical behavior of the joint as good as possible. To be able to solve the numerical instabilities of the timber joints, a deeper look needs to be done to the modelling of the wooden material and the steel wood contact. For this thesis a previously developed 3D numerical damage model of wood has been studied. This model has been elaborated by Sandhaas(2012) and it describes crack initiation and propagation of the material based on the concepts of continuum damage mechanics. The basic material model of wood has been implemented as a user material in the UMAT subroutine of ABAQUS. The developed model is giving some numerical instabilities due to the extreme distortion of the elements. During this thesis the model has been enhanced in order to be able to represent the mechanical behavior of wood as good as possible and solve the problem of the model. The modelling outcomes were compared to the results obtained by experimental tests (ref. to Sandaas,2012).The results showed that the first model, a tension test parallel-to grain, had been enhanced. Indeed the results got closer to the experimental value than the original model’s results did. The second model represented a timber joint with slotted-in steel plate with a dowel. The analysis were done with different wood spieces (spruce, beech and azobè). Regarding the spruce, the analysis reached fairly accurate results concerning the capacity load but they were less precise regarding the displacement and the stiffness. The prediction quality was rather poor for the other two species, beech and azobè. It is necessary to find other ways to further enhance the model.Even today a model that is able to represent all three fields (stiffness, capacity load and displacement) accurately doesn’t exist. Good results of one of these lead to bad results of the others. Modelling wood then still represents an evolving challenge.

The overarching objective in this work is to advance damage modelling for performance-based earthquake engineering. To achieve this objective, this thesis provides a new vision, technique, and software framework for the assessment of seismic damage and loss to building components. The advent of performance-based earthquake engineering placed a renewed emphasis on the assessment of damage and monetary loss in structural engineering. Assessment of seismic damage and loss for decision making entails two ingredients. First, models that predict the detailed damage to building components; second, a probabilistic framework that simulates damage and delivers the monetary loss for the reliability, risk, and optimization analysis. This motivates the contributions in this thesis, which are summarized in the following paragraphs. First, a literature review is conducted on models, techniques and experimental studies that address component damage due to earthquakes. The existing approaches for prediction of the seismic damage, repair actions, and costs are examined. The objective in this part is to establish a knowledge bank that facilitates the subsequent development of probabilistic models for seismic damage. Second, a logistic regression technique is employed for developing multivariate models that predict the probability of sustaining discrete damage states. It is demonstrated that the logistic regression remedies several shortcomings in univariate damage models, such as univariate fragility curves. The multivariate damage models are developed for reinforced concrete shear walls using experimental data. A search algorithm for model selection is included. It is found that inter-story drift and aspect ratio of walls are amongst the most influential parameters on the damage. Third, an object-oriented software framework for detailed simulation of visual damage is developed. The work builds on the existing software Rt. Emphasis is on the software framework, which facilitates detailed simulation of component behaviour, including visual damage. Information about visual damage allows the prediction of repair actions, which in turn improves our ability to predict the time and cost of repair.<br>Applied Science, Faculty of<br>Civil Engineering, Department of<br>Graduate

How DNA repair proteins search and recognise the rare sites of damage from the massive numbers of normal DNA remains poorly understood. FapydG (2,6-diamino-4-hydroxy-5-formamidopyrimidine) is one of the most prevalent guanine derived lesions involving opening of the imidazole ring. It is typically repaired by formamidopyrimidine-DNA glycosylase (Fpg) as an initial step in base excision repair; if not repaired, the lesion generates a G: C -+ T: A transversion. Unfortunately, studies on the recognition of FapydG have been hindered by difficulties to synthesise and incorporate the FapydG residue into a DNA duplex. Crystal structures of Fpg-DNA complexes have demonstrated three common recognition events: the protein specifically binding to the extrahelical lesion, bending DNA centred on the damaged base, and flipping the damage into the pocket. Thus, molecular modelling and dynamics simulation have been used to gather dynamical information of those recognition events for damaged and undamaged DNA. The simulations were initially performed when FapydG or G occurs in several dodecamer B-DNA sequences in aqueous solution, then inside the lesion-recognition pocket of Fpg, and during the flipping pathway from the helical stack to an extrahelical position. The influence of the damage on DNA stability and flexibility was first investigated. Energetic analysis revealed that damage to DNA does appear to destabilise the duplex. DNA curvature analysis and a novel combined method of the principal component analysis (PCA) and the Mahalanobis distance (DM) indicated that damaged DNA can adopt the observed protein-bound conformation with lower energetic penalties than its normal counterpart. Results of these studies have provided the validation of DNA bending enhancement by the FapydG lesion. It also suggested that intrinsic DNA bending could be a principal element of how the repair protein locates the lesion from vast expanse of normal bases. Considering the specific recognition of FapydG by Fpg, the aF-/39 loop of the Fpg enzyme may function as a gatekeeping to accommodate the lesion while denying the normal base. Remarkably fluctuating movement of the flipped G residue and the aF-ß9 loop is due to the formation of the non-specific Fpg/G complex with a lower binding energy by 8.4 kcal/mol compared to the specific Fpg/FapydG complex. Free-energy profiles for both damaged and undamaged base flipping were generated from the umbrella sampling simulations and the Weight Histogram Analysis Method (WHAM). An energy barrier for flipping the damage out from the helix is 2.7 kcal/mol higher than its equivalent G and the lesion is highly stabilised inside the pocket. In contrast, G flipping seems to be rapidly rotated out and into the duplex without the formation of a specific complex. These studies could unravel a potentially comprehensive process of the repair protein to find and recognise the lesion through the slow kinetic pathway in which the more deformable damaged DNA is initially located by the protein; the protein subsequently compresses the duplex into an appropriate angle and direction to form a specific protein-DNA complex prior to being flipped and repaired.

The behaviour of damage parameters in adhesive bonding has been investigated in order to predict the location of the initial crack in the adhesive region. Research started using bulk adhesive in which the loading is uniaxial, In this case the triaxiality, a variable that defines the stress state will have a value around unity.

The main objective of this research is to develop a numerical model for the analysis of frost damage to masonry and implement it in a program for 3-dimensional analyses of masonry structures using the finite element method. It is shown that the standard FE approach in analysing masonry structures is not economical, so a homogenization technique was used to substitute composite masonry material with an equivalent homogeneous one. Parametric studies of relationships of property between constituents and the equivalent material are presented together with validation of the procedure. Masonry, or its constituents, is treated as a brittle material with a tensile cut-of failure criterion. Again, in order to achieve more computational efficiency, homogenization technique is used to represent cracked material. A set of freeze-thaw tests are described which are important in detecting the most influential factors controlling frost durability of masonry. Models of frost action are investigated on a micro scale - scale of individual pores. A stochastic network model is developed with an aim to investigate permeability characteristics of the porous material. The results of these calculations are used in providing information necessary to calculate stress due to the frost action. Typical examples of frost damage to masonry are simulated showing good agreement with the expected results. The following conclusions are drawn: * The major cause of damage to masonry is the expansion of water on its phase change from a fluid to a solid phase in the 'sealed container' conditions. * Micro structural properties of the material are the most important single factor in deciding the amount of internal loading on the material matrix caused by ice formation. In order to calculate this loading simulation of the process on the micro scale is required. * It is confirmed that a high fraction of small pores in the porous structure causes higher level of frost susceptibility.

The aim of this research was to develop a reliable predictive fatigue damage model for adhesively bonded structures. It was necessary for such a numerical model to be independent of geometry of the structure and capable of considering different fatigue damage phases, of simulating the experimentally measured damage evolution and of predicting the effect of main fatigue loading characteristics. Three different adhesively bonded joints, namely the single lap joint, the laminated doubler in bending and the mixed-mode flexure specimen manufactured with the same adhesive system were considered for experimental and numerical investigations. The bonded joints were tested under quasi-static and fatigue loading and the failure responses of the bonded joints were studied experimentally and modelled numerically. To assess static and fatigue progressive damage in the bonded joints, experimental approaches, including backface strain and video microscopy techniques were employed. The effect of important fatigue loading parameters including the maximum fatigue load level and the load ratio on the failure behaviour of the bonded joints was examined experimentally. A cohesive zone model with a bi-linear traction-separation response was used to simulate the progressive damage in the adhesively bonded joints. This cohesive zone model was integrated with a damage mechanics based fatigue model to simulate the deleterious influence of fatigue loading. The proposed fatigue damage model was able to account for the effects of fatigue loading characteristics including the maximum fatigue load and fatigue load ratio. The static and fatigue damage models were calibrated, validated and optimised against the experimental results obtained and other published experimental data. The fatigue damage model was applied to adhesively bonded joints subjected to constant and variable amplitude fatigue loading. The model was able to successfully predict the detrimental effect of the variable amplitude fatigue loading as well as the constant amplitude fatigue loading. The proposed fatigue damage model was generally found to be a significant improvement on other damage models available for adhesively bonded structures

Hochwasserschadensprozesse werden von den drei Komponenten des Hochwasserrisikos bestimmt – der Gefahr, der Exposition und der Vulnerabilität. Dabei bleiben wichtige Einflussgrößen auf die Vulnerabilität, wie die private Hochwasservorsorge aufgrund fehlender quantitativer Informationen unberücksichtigt. Diese Arbeit entwickelt daher eine robuste statistische Methode zur Quantifizierung des Einflusses von privater Hochwasservorsorge auf die Reduzierung der Vulnerabilität von Haushalten bei Hochwasser. Es konnte gezeigt werden, dass in Deutschland private Hochwasservorsorgemaßnahmen den durchschnittlichen Hochwasserschaden pro Wohngebäude um 11.000 bis 15.000 Euro reduzieren. Hochwasserschadensmodelle mit Expertenwissen und datengestützten Methoden sind dabei am besten in der Lage Unterschiede in der Vulnerabilität durch private Hochwasservorsorge zu erkennen. Die über Hochwasserschadenprozesse erhobenen Daten und Modellannahmen sind von Unsicherheit geprägt und so sind auch Schätzungen mit. Die Bayesschen Modelle, die in dieser Arbeit entwickelt und angewandt werden, nutzen Annahmen über Schadensprozesse als Prior und empirische Daten zur Aktualisierung der Wahrscheinlischkeitsverteilungen. Die Modelle bieten Hochwasserschadensschätzungen als Verteilung, welche die Bandbreite der Variabilität der Schadensprozesse und die Unsicherheit der Modellannahmen abbilden. Hochwasserschadensmodelle, hinsichtlich der Prognoseerstellung und Anwendbarkeit. Ins Besondere verbessert die Verwendung einer Beta–Verteilung die Zuverlässigkeit der Modellergebnisse im Vergleich zu den häufig genutzten Gaußschen oder nicht parametrischen Verteilungen. Der hierarchische Bayessche Ansatz schafft eine verbesserte Parametrisierung von Wasserstand-Schadens-Funktionen und ersetzt so die Notwendigkeit empirischer Daten durch regional- und Ereignis-spezifisches Expertenwissen. Auf diese Weise kann die Vorhersage bei einer zeitlich und räumlichen Übertragung des Models verbessert werden.<br>Flood damage processes are influenced by the three components of flood risk - hazard, exposure and vulnerability. In comparison to hazard and exposure, the vulnerability component, though equally important is often generalized in many flood risk assessments by a simple depth-damage curve. Hence, this thesis developed a robust statistical method to quantify the role of private precaution in reducing flood vulnerability of households. In Germany, the role of private precaution was found to be very significant in reducing flood damage (11 - 15 thousand euros, per household). Also, flood loss models with structure, parameterization and choice of explanatory variables based on expert knowledge and data-driven methods were successful in capturing changes in vulnerability, which makes them suitable for future risk assessments. Due to significant uncertainty in the underlying data and model assumptions, flood loss models always carry uncertainty around their predictions. This thesis develops Bayesian approaches for flood loss modelling using assumptions regarding damage processes as priors and available empirical data as evidence for updating. Thus, these models provide flood loss predictions as a distribution, that potentially accounts for variability in damage processes and uncertainty in model assumptions. The models presented in this thesis are an improvement over the state-of-the-art flood loss models in terms of prediction capability and model applicability. In particular, the choice of the response (Beta) distribution improved the reliability of loss predictions compared to the popular Gaussian or non-parametric distributions; the Hierarchical Bayesian approach resulted in an improved parameterization of the common stage damage functions that replaces empirical data requirements with region and event-specific expert knowledge, thereby, enhancing its predictive capabilities during spatiotemporal transfer.

Ionising radiation is commonly used in radiation treatment of cancer to target cancerous cells. Upon entering a cell, ionising radiation can damage DNA directly or it can excite the cellular medium. The most abundant species produced following excitation of the medium is the low energy electron (LEE). LEE's can cause significant damage to DNA but the exact mechanism of this damage is unknown. This thesis investigates the indirect damage that LEEs do to DNA fragments, with particular emphasis on the influence of the surrounding water molecules. Using ab initio Molecular Dynamics, we investigated DNA-water systems, ranging from nucleobases to nucleotides. We found that water molecules in the environment are capable of forming hydrogen bonds with or transferring protons to the DNA fragment. In all these cases, when the DNA is H-bonded to or protonated by the water molecules, the energetic barriers to DNA damage reactions are affected. In the final parts of the thesis, we extended the research and investigated chemo-radiation therapy involving the chemotherapeutic agent cisplatin. We found once again that including the water environment around the DNA-cisplatin had an effect on the reactivity of this system with LEEs. Our results highlight the significant role that water molecules can play in DNA damage processes and demonstrate that it is therefore important to incorporate explicit water molecules in any simulation of the DNA damage process.

The current available damage models do not accurately predict effective plastic strain to failure in low triaxiality stress states. A damage model was developed for low triaxiality that is appropriate to hot rolling of steel. This work focuses on nucleation and growth of damage as well as the effect of the strain and stress path. The latter is especially important for the rolling of bar and other complex cross-section products. A study of damage mechanisms and methods to model them has been undertaken. It is pointed out that the many models are only useful under certain conditions but can be used when the expected damage mechanisms are active. Several test types were evaluated to assess their ability to simulate stress state in rolling. A program has been written to evaluate the stress state for plane and axisymmetric tests, which allows one to choose the most appropriate test-piece geometry. A test has been designed and implemented. Thermal and mechanical data was gathered, which has been used to relate the stress triaxiality to damage growth and identify appropriate damage growth models. The size and spacing distributions of inclusions in free cutting steels have been measured. The different distributions have an effect on the ductility of the different steels. This effect has been found to change at different strain rates and temperatures. By better accounting for the effect of inclusions on damage growth under a range of test conditions, the damage model can be significantly improved. Free cutting steels that contained different additions of heavy metals were tested. The ductility and damage mechanisms were compared in each of the steels. The effect of the precipitation of the different heavy metals at the inclusion to matrix boundary was highlighted. The same damage mechanisms were observed in each steel but the ability to accommodate damage varied between the steels. Ex-situ synchrotron x-ray micro-tomography was used to better measure and quantify the distribution of inclusions and damage evolution in a free cutting steel. Localised damage coalescence away from the centre of the uniaxial tensile test-piece was attributed to the effect of inclusion clustering. This research was used to develop a realistic damage model, which can predict damage growth and coalescence for a range of forming parameters and different stress-state conditions related to hot rolling applications. The micro-mechanics based model includes the effects of inclusion distribution on damage. The model is calibrated using twenty six temperature based material constants.

The classical theory of poroelasticity focuses on the coupled response of fluid flow and elastic deformation of porous media saturated with either an incompressible or a compressible pore fluid. The Theory of poroelasticity has been successfully applied to examine time-dependent transient phenomena in a variety of natural and synthetic materials, including geomaterials and biomaterials. The assumption of elastic behaviour of the porous skeleton in these developments is a significant limitation in the application of this theory to brittle geomaterials; which could exhibit non-elastic phenomena either in the form of initiation and extension of discrete fractures, or in the form of initiation and evolution of continuum damage in the porous skeleton. The computational methodology developed in this study examines the effect of development of such defects (fracture or damage) on the fluid transport characteristics and the poroelastic behaviour of saturated geomaterials. The finite element based computational models for fracture and damage phenomena examine two-dimensional plane strain and axisymmetric problems. The classical theory of linear elastic fracture mechanics is extended to examine the timedependent behaviour of local effects at the crack tip in poroelastic media. The numerical procedure accounts for the stress singularity of the effective stress field at the crack tip. The damage model based on the concept of continuum damage mechanics, takes into account the alteration of the stiffness and permeability characteristics of porous material due to development of micromechanical damage in the porous skeleton. The isotropic damage criteria governing the evolution of stiffness and permeability parameters are characterized by the dependency of damage parameters on the distortional strain invariant.<br>As the applications of the theory of poroelasticity diversify, attention needs to be focused on other aspects of importance. The class of transient and steady crack extension in poroelastic media is recognized as an area of interest in geomechanics applications and in energy resources recovery from geological formations. A computational algorithm is developed to examine the transient quasi-static crack extension in poroelastic media where the temporal and spatial variations of boundary conditions governing the displacement, traction and pore pressure fields are taken into account in the incremental analysis. The path of crack extension is established by a mixed-mode crack extension criterion applicable to the porous fabric. The computational modelling of steady state crack extension in poroelastic media at constant velocity is also examined for the plane strain problems. The finite element formulations of the governing equations, which are velocity-dependent, are developed by employing the Galerkin technique. The poroelastic behaviour of material depends on the propagation velocity at the crack tip. The computational schemes developed in this study followed an extensive procedure of verification via known analytical solutions to poroelasticity problems and for limiting cases of initial undrained (t &rarr; 0+) and final drained (t &rarr; +infinity) elastic responses recovered through analogous problems in classical elasticity.

Continuous casting is used to solidify most of the steel produced in the world every year. The process reduces the number of required milling stages and results in qualitative semi-finished products such as billets, blooms and slabs which will later be rolled into more specific shapes. Extending the range of finished product sizes produced from a given concast bloom or billet section is often limited by the minimum area reduction required to ensure effective consolidation and final mechanical properties. Predicting effective consolidation or level of remnant porosity has always been an important issue for steel producers as it will affect the mechanical properties of final products (strength, ductility, etc.). It is known that partial or complete recovery of strength in such porous materials can be obtained by pore closure and diffusive healing processes at elevated temperature. Devising an appropriate healing process which does not cause discontinuity in the microstructure and mechanical properties at the healing sites and prevents distortion of the component during bonding requires an accurate choice of thermo-mechanical processing parameters. Although there has been considerable work on materials such as titanium alloys, aluminium alloys and copper, damage healing in free cutting steel has not received much attention. The main aim of this research is to develop a realistic damage healing computational approach that can predict damage healing or recovery during soaking under different compressive stress levels, and be used for hot rolling applications. This study investigates the void elimination process through two stages of void closure and healing. An Abaqus/UMAT subroutine has been developed for the analysis of the material porosity elimination process including two stages of void closure and healing. This study uses the Gurson-Tvergaard model under hydrostatic compression to predict the void closure. A novel approach has been developed in the present work to identify the Gurson-Tvergaard model parameters using a non-gradient based optimisation search method (Pattern Search Method). The healing process is modelled based on a combination of diffusion bonding, creep and plasticity following the Pilling model and can be adapted to any other healing/diffusion bonding model. The material model has been calibrated for free cutting steel and a stress state representative of the rolling process, and used to predict the closure and healing processes under rolling. The effect of parameters such as Roll Gap shape Factor (RGF), initial amount and distribution of void volume fraction on porosity elimination during rolling has also been investigated. An experimental technique has been developed to identify the conditions (temperature, pressure, time) required for void elimination in Free Cutting Steel (FCS). Different combinations of load and time were tested and optimum conditions have been obtained. Tensile tests on the bonded specimens have been carried out to measure the strength of the bonded region. The position of fracture on the specimen and also the cross section of the fracture surface have been inspected. The experimental results have been used to calibrate the developed void elimination model. Using the developed model, predictions of densification and healing can be made for optimisation of the rolling schedule.

A fully coupled global-local approach for structural analysis has been developed. It is motivated by the need to use a range of scales and modelling techniques when designing a structure in composite materials. These range from the microscale at which the interfaces between fibres and matrix, or buckling of fibres themselves may play a role in the material behaviour, through intermediate scales where delamination and debonding may have an influence up to the macroscale where entire structures may be modelled with service loads directly applied. The method is based on passing boundary conditions from larger to smaller length scale models while passing information about damage and stiffness degradation up through the scales. By using nested levels of submodel, a greater range of length scales may be included in a single set of coupled analyses. Here an explanation of the methods of coupling two scales of solid models as well as coarse shell models to relatively refined solid models is presented. Each of these methods is validated against equivalent models using established modelling techniques, and are shown to produce results comparable to a complete model at the refined scale and preferable to other global-local approaches. Experimental tests have also been carried out on a stiffened panel with two stiffener runouts undergoing debonding. Not only did the coupling method model these tests accurately, but it was also shown to be more appropriate than simple submodelling in this case. A further demonstration of the techniques is included. The largest scale consisting of a shell element mesh is coupled with an intermediate scale with a continuum shell mesh, which in turn is coupled to a refined scale solid model. This demonstration shows how the methods developed here could be used to unify various analyses in the composites design process which until now have remained separate.

The goal of this study is to gain insight into mechanisms of soil-structure interaction for buildings affected by tunnel excavation and to discuss reliable methods to evaluate the potential damage. Assessing the structural response to excavation-induced deformations involve a combination of both geotechnical and structural aspects, such as soil behaviour, building behavior, soil-structure interaction and modelling techniques. This study focuses on the behavior of reinforced concrete buildings founded on coarse-grained soils. In particular, reference was made to the case study of new Milan (Italy) metro-line 5, excavated in sandy soils by an earth pressure balance (EPB) machine. Observed volume loss was averagely equal to 0.5%, which did not produce any damage on surrounding buildings due to efficient excavation procedure. The goal of this work is to simulate different scenarios of volume losses and structural configurations through numerical FE simulation. Therefore, a maximum volume loss of 3%, index of inappropriate execution of excavation work was considered. Moreover, the influence of structural stiffness is investigated by taking into account different building configurations including the presence of infills and openings due to doors and windows. A 5-storey RC building founded on a strip footing and interacting with tunnel excavation is modelled in detail. The numerical analyses are performed with the commercial software Abaqus. Damage assessment is carried out both by direct analysis of the stress-strain response of the structure in the numerical analyses and by empirical and analytical methods typically used in the engineering design. Recent researches illustrate that infill masonry walls should be taken into consideration as far as they are most sensitive elements to damage when affected by excavation. This thesis illustrates the beneficial effect, in reducing the deflection ratio and damage to building when modeling it with presence of infills.

To date, there are no international codes or standards that deal with estimation offatigue in subsea wellhead systems. Nevertheless have preliminary analytical methodsfor wellhead fatigue estimation been established. These analytical methods involve theuse of global dynamic response analyses. Such analyses are commonly carried out innite element software where the drilling system is modelled as beam or bar elements.Several uncertainties exist in regards to the mathematical modelling and simulationin global response analyses of drilling systems. In this thesis the uncertainties thatare related to the goodness of the representation of the blowout preventer stack (BOPstack), are addressed.An overview of previous and ongoing work on analytical estimation of wellhead fatigue isgiven. Relevant theory on the subject is presented and described. The theory comprisesof fatigue on structures, loads on a drilling system and static and dynamic responseanalysis. The main features of the preliminary analytical methods for estimating wellhead fatigue are summarized.The BOP stack is commonly assumed to have innitely high stiness when performinga global response analysis of a drilling system. The main objective in the thesis is toinvestigate if this is a good assumption. The investigation start with local modellingof an elastic beam element model of a BOP stack. Further, this elastic beam model iscalibrated to have the same stiness properties as a detailed 3D element model of theBOP stack. The calibrated elastic beam BOP stack model is implemented in a globalmodel of a drilling system. Global response analyses are carried out for two globalmodels. One with an innitely sti BOP stack model and one with the calibratedelastic BOP stack model. Fatigue damage estimates are calculated with basis in theresults from the global response analyses. The eects of BOP stack modelling areevaluated with regards to estimated fatigue damage in the wellhead. In addition to this main study, parameter studies and a sensitivity study are carried out to evaluateuncertainties and assumptions within a realistic frame.The results from the wellhead fatigue assessments conclude that the elastic BOP stackmodel imposes greater estimated fatigue damage in the wellhead compared to theinnitely sti BOP stack model. The dierence, in terms of estimated fatigue damage,imposed by the two BOP stack models is, at maximum, 0.51% for the main study. Thelargest dierence observed in the parameter studies is 1.34%. Hence, it is concluded thatthe eects of improved BOP stack modelling in a global response analysis, with respectto wellhead fatigue estimation, is not signicant. There are though some uncertaintiesconnected to the bending stiness of the wellhead connector and the LMRP connector.

Accurate assessment of the damage to buildings due to tunnelling in soft ground becomes an important issue when a tunnel is constructed under historic masonry buildings in urban areas. The current two-stage approach in which settlements estimated from a Gaussian curve are applied to a building does not consider soil-structure interaction and fails to give a correct prediction of the damage. This thesis describes a complete threedimensional finite element model for assessment of the settlement damage to masonry buildings induced by tunnelling in London Clay, and an investigation of the interaction between a masonry building and the ground. A macroscopic elastic no tension model, which assumes the material has zero tensile strength but infinite compressive strength, is developed to simulate the behaviour of masonry. Numerical techniques are proposed to improve the stability of the calculation. The comparison of the no tension and elastic models, by applying Gaussian curve settlement troughs to both a plain wall and a facade, shows that the no tension model predicts different behaviour of the masonry building during tunnelling, including different cracking patterns and damage grades. Two-dimensional finite element analyses combining the building, modelled by the no tension material, and the ground, modelled by a nested yield surface model, give insight into the interaction between the masonry structure and the ground. They suggest the importance of the stresses in the soil prior to the excavation in affecting the ground movements during tunnelling. Thus the weight of the building controls the overall magnitude of the ground movements beneath the building, while the stiffness of the building affects the shape of the trough. A key aspect of the behaviour of the masonry building is the formation of stress arches. Finally the three-dimensional finite element analyses are described. Both symmetric and unsymmetric cases are analysed. The results show that the three-dimensional analysis gives more realistic modelling of the problem and is likely to be necessary for practical situations, especially when a building is not symmetrically located with respect to the tunnel - a case which cannot be analysed in two-dimensions. A special tying scheme is proposed for the connection of the nodes belonging to elements of different types, which are defined in their own local co-ordinate systems. Different types of tie elements are formulated and implemented for connection between two-dimensional and three-dimensional elements in various combinations.

PONTIFÍCIA UNIVERSIDADE CATÓLICA DO RIO DE JANEIRO<br>Após identificar que uma trinca de fadiga permanecia fechada durante parte do ciclo, Elber assumiu que o dano era induzido apenas pela fração do carregamento acima da carga necessária para abrir a trinca. Diversos modelos foram propostos utilizando o Delta Keff como força motriz da propagação, como os modelos da faixa plástica (strip-yield), que são amplamente utilizados para prever vida residual de componentes trincados. Embora o fenômeno do fechamento da trinca esteja provado, sua real importância na propagação da trinca de fadiga ainda é controversa. Outros mecanismos, além do fechamento da trinca, foram utilizados na tentativa de explicar os efeitos de sequência do carregamento na propagação em amplitude variável como o campo de tensão residual à frente da trinca. Mesmo após mais de 50 anos de pesquisas desde a proposição da primeira regra de propagação por Paris ainda não há consenso nem sobre o mecanismo nem sobre a modelagem. Esse trabalho tem como objetivo apresentar uma modelagem para prever propagação da trinca de fadiga com base na hipótese de que o dano acumulado por deformação plástica seria a força motriz para propagação. A modelagem proposta se diferença de outros modelos de acúmulo de dano por permitir que o contato existente entre as superfícies da trinca exerça influência sobre as deformações plástica à frente de sua ponta. Os resultados mostram que a modelagem proposta possui capacidade de reproduzir curvas de propagação semelhante ao modelo strip-yield.<br>After identify that a fatigue crack remains closed during part of the load cycle, Elber assumed the damage was induced only by the cycle part over the load required to open the crack. Several models were developed based on Delta Keff as the strip-yield ones, which are widely used to predict residual lives of cracked components. Although the crack closure phenomenon is well proven its actual significance for the propagation is still controversial. Others mechanisms, beyond the crack closure, were used in trying to explain the sequence effects on variable amplitude crack propagation like the residual stress field ahead of the crack tip. However even after more than 50 years of research since the first propagation rule proposed by Paris there is no neither about the mechanism neither about modelling. This work has the aim of present a modelling to predict fatigue crack growth based on the hypothesis that the damage accumulated by cyclic plastic strain would be propagation the drive force. The modelling proposed differs from others damage accumulation models by allowing the existed contact between the crack surfaces to exercise its influence on plastic strain ahead of the crack tip. The results show that the proposed model is able to reproduce propagation curves similar to the model strip-yield.

A computational study of the size effects of open-hole tension composite laminates is carried out. The thickness-dependence of translaminar fracture toughness is accounted for in the numerical model, which enables the sublaminate-scaling effect of strength to be accurately predicted by a deterministic model. Neglecting delamination in modelling is found to cause mesh-dependence and over-estimation in strength predictions. A smeared crack model with cohesive elements between plies can reliably predict the failure mode, but not the strength, for laminates failed by delamination. A floating node method is developed for explicitly modelling multiple discontinuities within an element. The degree-of-freedom vectors do not have associated initial coordinates; they are assigned to new material points when needed during analysis. The proposed method is well suited for modelling strong, weak and cohesive discontinuities, for the representation of complex crack networks, and for the accurate modelling of matrix crack/delamination interactions in composites.

In this thesis we use a time-dependent tight-binding model metal evolving under semiclassical Ehrenfest dynamics to explore the effects of electron-ion energy exchange on radiation damage phenomena. By incorporating an explicit model of quantum mechanical electrons coupled to a set of classical ions, our model correctly reproduces the interaction of excited ions with cooler electrons and captures phenomena absent in classical molecular dynamics simulations and in much-used analytical models. With our simple model we have been able to simulate large numbers of radiation damage cascades. We have directly explored the electronic excitations stimulated in such cascades and have found them to be well characterized by an elevated electronic temperature. We have also analysed the effect of these excitations in weakening the bonding interactions in our model metal, and the effect of these weakened interactions on the evolution of replacement collision sequences. By separating out components of the Hellmann-Feynman forces exerted by the electrons on the ions, we have identi ed the non-adiabatic force, resulting from the finite response time of the electrons to ionic motion and responsible for the accumulating electronic excitations. Based on simplifying physical arguments we have derived a temporallyand spatially-local expression for this force suitable for incorporation within a classical MD code at very low computational cost. Data from our simulations show that our new expression for the non-adiabatic force captures much of the microscopic detail of the direction and magnitude of the force. We find that it significantly outperforms commonly used viscous damping models of ion-electron energy transfer. At higher energies, our simulations of ion channelling reveal a new resonant enhancement of the electronic charge on the channelling ion and corresponding effects on the stopping force. We explain these phenomena with reference to the detailed atomic and electronic structure of our model.

Among the bulk metal forming processes hot forging is often the only option if large reductions of the forging load are required or if the material formability needs to be significantly increased or even if specific thermally-induced microstructural changes are needed to take place during the deformation process. Moreover it still retains at some extent the positive features of the cold forging processes: high production rates, complex final shapes attainable and little to no material waste. Hot forging is therefore used for the production of large parts, with complex shapes and especially when using materials characterized by low formability and high toughness at room temperature or if particular microstructural characteristics are required. It is worth to be underlined that these are often the features of the innovative metal alloys that have been increasingly being used in the last decade and namely: Mg alloys, Al alloys, Ti alloys and superalloys. Finite Element metal forming numerical simulation has become an increasingly important process optimization tool, due to the growing computational power available at reduced costs, which spread it in the industrial world. Its use allows reducing process design time and prototyping costs as well as long and expensive plant downtimes for process variable tuning. For all these reasons hot forging has become a strategic process and its accurate numerical simulation is encountering great industrial interest. One of its main targets is the determination of the maximum strain that the material can undergo during the deformation process, since it is strictly related with both the final shape and the surface quality of the manufactured part. In this sense hot formability modelling provides a meaningful example of a topic which is both of great scientific and industrial interest. From the scientific point of view ductile damage modelling has been developed at first for cold processes, for which crack formation is usually a major issue. Last decade research efforts focused on the development of more advanced fracture criteria, keeping into account a complete characterization of the stress state dependency of formability. However the complex analytical formulation of these criteria and their expensive experimental calibration kept them de facto away from practical industrial use. On the other hand hot formability modelling was traditionally addressed by direct application of the conventional cold fracture criteria, under the implicit assumption of isothermal conditions. This approach has evident limits since it does not account for temperature influence on material formability and it does not even provide any physical insight on the different fracture mechanisms that can develop as temperature changes. In recent years some research efforts have been performed on a deeper investigation upon this point: experimental formability campaigns on different metal alloys were carried out in order to assess the temperature and strain rate influence, while some analytical models were proposed in order to describe the damage evolution at high temperature and in particular the onset of the so-called “hot shortness”. However these models are still quite simple and can describe only a limited variety of formability trends, losing in accuracy once complex microstructural phenomena take place. Moreover their validation has been done with simple laboratory tests and not with real processes industrial trials in which non-uniform thermo-mechanical conditions are present and the material can evolve through fracture mode-changing regimes. The objective of this work is the elaboration of a novel approach to hot ductile fracture modelling, capable to represent accurately the formability evolution of a metal alloy as a function of both the main thermo-mechanical variables and its physical and microstructural characteristics, yet providing a tool simple enough to be of industrial utility. To this aim the hot cross wedge rolling of a precipitation hardened aluminium alloy was taken as industrial reference case, since it is an innovative and non-standard forging process which entails variable and non-homogeneous thermo-mechanical conditions. The study case is of remarkable interest since it has a narrow process temperature window, being limited at the top by Mannesmann-type axial cracking onset and at the bottom by unwanted grain coarsening. Moreover the metal alloy used, the AA6082-T6, has the microstructural features, namely the intermetallic precipitates, that make it a good example of a metal alloy of wide industrial use, that during forming can undergo complex microstructural changes. A hot tensile test campaign was performed on a wide range of thermo-mechanical conditions and the results highlighted an unexpected negative strain rate influence on formability. Fractographic and micro-chemical analysis were then performed in order to assess the microstructural reasons of this behaviour and finally two approaches to the material formability modelling were proposed and calibrated extending to hot conditions the classic Oyane-Sato fracture criterion. The first one consists on the empirical calibration of the criterion by means of a bi-linear interpolation of the experimental data, while the second one entails a physically-based analytical formulation of the material fracture locus, which has also the advantage of being of easier calibration. These models were then validated on the cross wedge rolling process simulation by comparison with the industrial trials results and the outcomes were critically assessed.<br>Tra i processi di formatura massivi di materiali metallici, la forgiatura a caldo è spesso l'unica opzione nei casi in cui siano richieste forti riduzioni del carico di forgiatura o se la formabilità del materiale deve essere notevolmente aumentata o anche se risulta necessario ottenere durante il processo di deformazione determinate modifiche microstrutturali indotte termicamente. Inoltre essa conserva ancora in certa misura le caratteristiche positive dei processi di forgiatura a freddo: alta produttività, possibilità di realizzare forme finali complesse e scarto di materiale ridotto o inesistente. La forgiatura a caldo viene quindi utilizzata per la produzione di pezzi di grandi dimensioni, di forma complessa e soprattutto quando si utilizzano materiali caratterizzati da bassa formabilità ed elevata tenacità o se particolari caratteristiche microstrutturali sono richieste. Vale la pena di sottolineare che queste sono spesso le caratteristiche delle leghe metalliche innovative che sono state sempre più utilizzate negli ultimi dieci anni e precisamente: leghe di magnesio, alluminio, titanio e superleghe. La simulazione numerica agli Elementi Finiti di processi di formatura di materiali metallici è diventata nell’ultimo decennio uno strumento sempre più importante per l’ottimizzazione di processo, grazie alla maggiore potenza di calcolo disponibile a costi ridotti, che ha permesso la sua diffusione nel mondo industriale. Il suo utilizzo permette di ridurre i tempi di progettazione del processo ed i costi di prototipazione ed anche lunghi e costosi tempi di fermo impianto per la taratura delle variabili di processo. Per tutti questi motivi la forgiatura a caldo è diventata un processo strategico e la sua accurata simulazione numerica incontra grande interesse industriale. Uno dei suoi obiettivi principali è la determinazione della massima deformazione che il materiale può subire durante il processo deformativo, dal momento che essa è strettamente legata sia con la forma finale che con la qualità superficiale del componente prodotto. In questo senso la modellazione della formabilità a caldo fornisce un esempio significativo di un argomento che è al tempo stesso di grande interesse scientifico e industriale. Dal punto di vista scientifico la modellazione del danneggiamento duttile dei materiali è stata originariamente sviluppata per le lavorazioni a freddo, per cui la possibile formazione di cricche è un problema di maggior rilievo. Gli sforzi di ricerca nell'ultimo decennio si sono concentrati sullo sviluppo di criteri di frattura più avanzati, che hanno permesso la caratterizzazione completa dell’influenza dello stato tensionale sulla formabilità. Tuttavia la complessa formulazione analitica di questi criteri e la loro costosa calibrazione sperimentale ne hanno di fatto impedito la diffusione in ambito industriale. D'altra parte la modellazione della formabilità a caldo è stata tradizionalmente affrontata tramite la diretta applicazione dei convenzionali criteri di frattura a freddo, sotto l'ipotesi implicita di condizioni isoterme. Questo approccio presenta evidenti limiti, in quanto non tiene conto dell'influenza della temperatura sulla formabilità del materiale, né permette di intuire il senso fisico dei diversi meccanismi di frattura che possono svilupparsi al variare della temperatura. Negli ultimi anni alcuni sforzi si sono fatti per approfondire quest’ultimo punto: campagne sperimentali di formabilità su diverse leghe metalliche sono state eseguite per valutare l'influenza della temperatura e della velocità di deformazione, mentre alcuni modelli analitici sono stati proposti per descrivere l’evoluzione del danneggiamento ad alte temperature ed in particolare l’insorgere della "fragilità a caldo". Tuttavia questi modelli sono ancora abbastanza semplici e possono descrivere solo una varietà limitata di comportamenti del materiale, perdendo in precisione nel caso avvengano fenomeni microstrutturali complessi. Inoltre la loro validazione è stata effettuata con semplici test di laboratorio e non su reali processi industriali in cui si sviluppano condizioni termo-meccaniche non uniformi e il materiale può evolvere attraverso regimi in cui i meccanismi di frattura sono variabili. L'obiettivo di questo lavoro è l'elaborazione di un nuovo approccio alla modellazione della frattura duttile a caldo, in grado di rappresentare accuratamente l'evoluzione della formabilità di una lega metallica come funzione sia delle principali variabili termo-meccaniche che delle sue caratteristiche fisiche e microstrutturali, restando al contempo uno strumento sufficientemente semplice da essere di utilità industriale. Per questo scopo è stata presa come caso di riferimento industriale la rullatura trasversale a caldo di una lega di alluminio indurita per precipitazione, dal momento che si tratta di un processo di forgiatura non convenzionale ed innovativo e che comporta condizioni termo-meccaniche variabili e non omogenee. Il caso di studio è di notevole interesse poiché è caratterizzato da una stretta finestra di temperatura di processo, limitata superiormente dall’insorgenza di criccatura assiale per effetto Mannesmann ed inferiormente da un indesiderato ingrossamento della grana cristallina. Inoltre, la lega metallica utilizzata, l'AA6082-T6, ha delle caratteristiche microstrutturali, vale a dire i precipitati intermetallici, che lo rendono un buon esempio di una lega metallica di largo uso industriale, che durante la formatura può subire variazioni microstrutturali complesse. Una campagna di prove di trazione a caldo è stata eseguita su un ampio spettro di condizioni termo-meccaniche ed i risultati hanno evidenziato un’inattesa influenza negativa della velocità di deformazione sulla formabilità. Analisi frattografiche e micro-chimiche sono quindi state eseguite al fine di valutare le ragioni microstrutturali di questo comportamento ed infine due approcci alla modellazione della formabilità della lega sono stati proposti e calibrati estendendo alle alte temperature il classico criterio di frattura di Oyane-Sato. Il primo consiste nella calibrazione empirica del criterio mediante interpolazione bi-lineare dei dati sperimentali, mentre il secondo si basa su una formulazione analitica physically-based del fracture locus del materiale, che ha anche il vantaggio di essere di più facile calibrazione. I modelli sono stati poi validati sulla simulazione del processo di rullatura trasversale comparandone i risultati con quelli delle prove industriali e valutandoli in modo critico.

Advanced material models for the intra- and inter-laminar damage and failure prediction of biaxial carbon/epoxy Non Crimp Fabric (NCF) composites are presented, which enable application to large scale practical composite structures for the automotive industry. The model established for intra-laminar failure combines the elasto-plastic continuum damage constitutive model first proposed by Ladevèze, with the matrix failure model of Puck, and improves the shear damage representation by using an exponential damage evolution law. An extensive test program has been conducted using biaxial NCF composites with differing degrees of fabric pre-shear. The presheared biaxial NCF are considered to take into account potential changes of the damage and failure properties occurring due to the draping process in real structures. The database established is used to identify parameters for the proposed Ladevèze-Puck matrix damage and failure model, allowing prediction of the main matrix failure modes of unsheared and presheared biaxial NCF composites. Furthermore, the failure prediction within zones of stress concentrations is addressed utilising flat NCF laminates with different holes and lay-ups. For numerical elastic failure prediction of these tests a simple fibre criterion is presented which is related to a characteristic element size. The Puck failure model is adopted to represent the inter-laminar (delamination) crack initiation strength of the investigated NCF. This stress based model for delamination crack initiation operates in conjunction with existing methods for the simulation of delamination crack propagation, which are related to the critical strain energy release rate. All material models that have been combined and further developed during the course of this work have been implemented in a research version of a commercial crashworthiness Finite Element code. The intra-laminar failure model is validated against a test series of transversely loaded circular- composite discs having differing degrees of fabric pre-shear. The final validation example for intra-laminar failure has considered a complex composite door structure subjected to lateral punch intrusion. The enhanced delamination failure model is validated using a composite beam structure, exhibiting delamination failure under local transverse tension loading. For all validation tests a good agreement between simulations and experiments has been demonstrated. It is believed by the author that the new contributions presented in this thesis significantly raise the level of numerical prediction of deformation, damage, and failure for Continuous Carbon Fibre Reinforced Composites. Special emphasis is placed on the fact that all material model parameters have been obtained from simple specimen tests, and have been used without adjustment for the simulation of subsequent Validation tests utilising Composite structures of high complexity. This continuous approach is in contrast to the majority of similar investigations found in the literature, in which the adjustment of material model parameters appeared to be necessary in order to achieve a good agreement between simulations and experiments of complex structures.

Composite materials have been increasingly used in the past two decades since they offer significant potential weight reduction, part design flexibility and improved specific mechanical performance compared to traditional metals. For specific applications, braid reinforced composites offer better near net shape part and manufacturing flexibility than conventional unidirectional laminates, albeit at the expense of slightly lower in-plane stiffness and strength. Furthermore, for impact and crash applications, which is the emphasis of this thesis, their tow waviness and interlocking can offer excellent damage tolerance and energy absorption. In this work, heavy tow (24k) biaxial carbon and glass braided preforms were used to manufacture coupons and beam structures to undertake an extensive testing campaign to characterise different damage and failure mechanisms occurring in braided composites. Due to large shear deformation and surface degradation, non conventional measurement techniques based on marker tracking and Digital Image Correlation were successfully used to measure strains in the damaging material. The modelling of braided composites was conducted using the meso-scale damage approach first proposed by P. Ladevèze for unidirectional composites. The calibration of an equivalent braid unidirectional ply was achieved using the experimental results obtained for different braided coupons. Furthermore, failure mechanisms observed experimentally, such as tow stretching and fibre re-orientation occurring during loading history, were integrated into the model. A new unidirectional ply formulation was subsequently implemented into the explicit finite element code PAM-CRASHTM. Validation of the new model using single element, coupons and beams were conducted that provided a satisfying correlation between experimental tests and numerical predictions.

Due to the complex nature of fibre reinforced composite (FRC) materials, the onset of damage does not cause instantaneous failure of the entire structure. As a natural progress of the research in the area of modelling damage at multi-scales, a discrete element method (DEM) has been presented in this thesis to simulate and analyse the damage progression in FRC laminates which consists of interfacial debonding, transverse cracking, delamination, and transverse cracking and delamination under quasi-static tensile and/or thermalloadings. Regular and random packing schemes have been developed to assemble the particles to construct the DEM models of the composite materials in which two contact constitutive models, i.e. parallel bond model and contact softening model, are used to represent the linear elastic properties of the fibre and the plastic or cohesive behaviour of matrix and interface, respectively. The interlaminar delamination in composite laminates under mode I, mode II and mixed mode have been modelled and the extension of plastic zone in front of the crack tip has been predicted. The initiation and propagation of matrix cracking as well as the consequent fibre/matrix interfacial debonding process in single-fibre composite has been modelled and analysed by DEM. The effects of fibre distribution and fibre volume fraction on the transverse cracking path and the residual damage remaining in the composite lamina are studied by DEM. The progression of transverse cracking and delamination in cross-ply laminates, the cracking density as well as the stiffness degradation have been predicted by DEM and compared with those from other numerical models or experiments. A thermal expansion scheme has been developed in the DEM modelling of composite laminates, and the damage progression of matrix cracking, fibre/matrix debonding, transverse' cracking and delamination are included in the thermal-mechanical coupling DEM model with the consideration of material microstructures. The outcome of this research has validated the application of DEM in composite laminates in terms of its advantages in the modelling of damage progression and the prediction of cracking density and stiffness reduction, and also proved the potential of DE M in the future research of composite material design.

Most of the detrimental effects of ionising radiation in biology, such as cell killing and mutation, result from double strand break in DNA. To help in understanding their mechanism of formation, biophysical modelling of DNA damage at a sub-cellular level has been carried out. A Monte Carlo program was written in Fortran 77 to simulate the exposure of a cellular system (biological structure model) to ionising radiation. The model evaluates the physical effect of the ionisation and excitation processes in three stages: (1) simulating the physical data using an event-by-event Monte Carlo track structure code TRAX, (2) calculating the distances between these events and their distribution, and (3) assigning reaction probabilities to the parameters in the survival model. The objective of this work is to highlight the important factors that should be included in any model of cell survival, namely the degree of detail required for the DNA and the track structure. For this purpose, two models of energy deposition were considered (random independent hits and structured tracks) combined with two models of DNA (homogeneously-distributed and “nucleosome cyclinder”). It was found that the homogeneous DNA model was not able to give good agreement with measured cell survival data, using either random hits or hits distributed along tracks. On the other hand, the nucleosome DNA model with the correct track structure provided a better fit to a family of experimental survival curves over a range of dose rates. Therefore it is essential to consider some structure to the DNA, together with a correct model for the tracks. In addition the combination of the nucleosome cylinder DNA model with track structure gave results that were compatible with experimental estimates of the numbers of single-strand and double-strand breaks per cell per Gy, and the ratio between these quantities.

The development and implementation of a complex numerical model for the determination of the damage to masonry buildings resulting from tunnelling settlements is described in this thesis. The current methods of damage prediction do not, in general, take into account the stiffness and weight of the surface structure. The model addresses this deficiency by explicit inclusion of the structure. Three-dimensional finite elements are used to model the ground with a non-linear, elasto-plastic soil model based on kinematic hardening. Tunnel linings are modelled using a novel overlapping elastic shell element: volume loss being simulated by shrinkage of linings coincidentally with excavation. Structures are modelled as collections of facades comprised of plane stress elements using a non-linear material model for masonry, similar to elastic-no tension. In developing the three-dimensional model, its two-dimensional counterpart is also studied. While the beam and shell elements used for linings (in two- and three-dimensions respectively) have the advantage of no rotational degrees of freedom the need to model boundary conditions at the element stiffness level complicates implementation. Tests using the shell elements show them to be satisfactory for the purpose of modelling tunnel linings. Results from a small number of analyses are given for construction of a straight tunnel beneath simple masonry structures. It is shown that the effect of the building on settlements depends heavily on its location in plan with respect to the tunnel axis. Predictions of crack patterns using the model for these analyses show that facades which the tunnel passes under first are less damaged than those later in the excavation sequence. Both of these conclusions serve to demonstrate that the problem can only be realistically modelled using three-dimensional methods. At present, however, the computer resources required to run the three-dimensional model are considerable.

In this thesis the influence of stress-state on ductile damage in free cutting steel under hot forming conditions is examined. The industrial motivation for the project focuses on edge cracking in hot rolling. A brief outline of the hot rolling process conditions is necessary to define the important parameters affecting edge cracking. Triaxiality was, thus, identified as the key parameter relating to damage under hot rolling conditions. Based on a detailed literature review, the appropriate testing and modelling methodology were identified for this body of work. A high temperature, uniaxial tension test program was implemented to identify the effect of triaxiality on damage under hot forming conditions. Double notched bars with varying notch radii were utilised, thus inducing different stress triaxialities due to geometrical constraints. Based on the resultant stress-strain data the effect of triaxiality on ductility and the strain to failure was investigated. Subsequently, unbroken notches from tested double notched samples were sectioned and optically examined to reveal damage initiation sites. Interesting damage features were identified and correlated with sample geometry (i.e. triaxiality) and testing conditions. Finite element analysis of the double notched samples revealed the effect of triaxiality on the local stress-state. The accuracy of the mechanical analysis from such simulations was improved by incorporating the thermal gradients induced during high temperature Gleeble tests. Three stress parameters were examined in relation to their effect on the experimentally observed damage; maximum principal stress, effective stress and hydrostatic stress. The maximum principal stress and equivalent stress were most clearly correlated to damage development under multiaxial conditions for this particular free cutting steel. Based on the results of the stress-state investigation of the double notched samples, a multiaxial damage expression was developed that reproduced the experimentally observed damage characteristics. The new multiaxial damage model was calibrated using a combination of uniaxial and multiaxial stress-strain data and damage profiles. The model was shown to have good accuracy in predicting both the stress-strain data and the damage initiation sites as a function of geometry and damage conditions. Finally, an extensive range of temperature and strain rate conditions were simulated for all tested sample geometries, and an additional sample geometry, to fully understand how testing conditions affect damage characteristics and under what triaxialities this is prone to happen.

Composite materials play an ever increasing role in the design of modern day aeronautical and automotive structures due to their weight saving potential. Generally progress in constituent material production and composite manufacturing have resulted in lower costs for composite structures, which has made them more attractive for a number of industries, including the aeronautical and automotive industries. However, while sufficiently accurate numerical models exist to model damage initiation and progression in metal structures similar models are not yet available for composite structures. Yet the ability to model damage accurately is an integral part of the design process in both the aeronautical as well as the automotive industry. Due to the more complex microstructure of textile composites compared to metals a numerical model to predict the behaviour of a macrostructure needs to take microstructural effects into account. Multi-scale modelling approaches are uniquely suited to efficiently incorporate not only micro-scale affects but also higher scale affects like tow buckling. Therefore a multi-scale approach to model damage initiation and progression in textile composites based on the finite element method is presented in this thesis. A number of mechanical tests of a benchmark composite are conducted to measure input parameters for the multi-scale approach as well as mechanical behaviour for comparison with model predictions. The multi-scale approach is used to predict the mechanical behaviour of the benchmark composite for two different load cases, pure tension and pure shear. Results for the pure shear load case show significant deviations between predicted and experimentally measured stress-strain curve. For the pure tension load case transverse strain predictions also deviate significantly from the experimental data, stress-strain data in the loading direction however show good agreement between predicted values and experimentally measured data. Whilst further improvements are still required, the approach presented in this thesis provides a solid foundation for designers to predict damage initiation behaviour and progression in textile composites.

This thesis investigates the use of the acoustic emission (AE) monitoring technique for use in identifying the damage mechanisms present in paper associated with its production process. The microscopic structure of paper consists of a random mesh of paper fibres connected by hydrogen bonds. This implies the existence of the two damage mechanisms of interest, the failure of a fibre/fibre bond and the failure of a fibre. The majority of this work focuses on the development of a novel hybrid mathematical model which couples the mechanics of the mass/spring model to the acoustic wave propagation model for use in generating the acoustic signal emitted by complex structures of paper fibres under strain. A discussion of the coupling method is presented and the model is then analysed using a simple plucked fibre as a test case with a comparison between the numerical and experimental results. The hybrid mathematical model is then used to simulate small fibre networks aimed at providing information on the acoustic response of each damage mechanism. To do this the mass/spring model must successfully simulate the response of the fibre structure when undergoing a fibre/fibre bond failure or a fibre failure. This can be achieved by dynamically manipulating the mass and spring elements of the fibre structure. The simulated AEs from the two damage mechanisms are then analysed using a Continuous Wavelet Transform (CWT) to provide a two dimensional time/frequency representation of the signal. From the CWT certain features of the AEs can be attributed to each damage mechanism and as such a criteria for the time and frequency properties of each damage mechanism can be formulated. This criterion provides the basis for identifying the damage mechanisms present in the experimental data. The final contribution of this thesis is the investigation of training an intelligent classifier which can dynamically identify the AEs from the two damage mechanisms. This is achieved by converting the time and frequency criteria for each damage mechanisms into a set of features for the training of a Self-Organising Map (SOM). The significant step in this analysis is the method for the extraction of the features from the CWT of the AE. This work successfully combines four different scientific areas, paper physics, acoustic emission technology, data analysis and computational modelling to provide an insight into the micro-mechanics of paper. The most significant contribution of this work is the development of the hybrid model which has the ability to generate the acoustic response of a paper fibre structure undergoing two different damage processes. This alone has provided a significant insight into the micro-mechanics of paper to allow for the identification of the two damage mechanisms when the AEs are analysed with the CWT. Other contributions include the method used for the extraction of relevant features from the CWT to enable the training of a SOM for identifying the type of damage mechanism the AE originated from.

Previous investigations relating to lightning strike damage of Carbon Fibre Composites (CFC), have assumed that the energy input from a lightning strike is caused by the resistive (Joule) heating due to the current injection and the thermal heat ux from the plasma channel. Inherent within this statement, is the assumption that CFCs can be regarded as a perfect resistor. The validity of such an assumption has been experimentally investigated within this thesis. This experimental study has concluded that a typical quasi-isotropic CFC panel can be treated as a perfect resistor up to a frequency of at least 10kHz. By considering the frequency components within a lightning strike current impulse, it is evident that the current impulse leads predominately to Joule heating. This thesis has experimentally investigated the damage caused to samples of CFC, due to the different current impulse components, which make up a lightning strike. The results from this experiment have shown that the observed damage on the surface is different for each of the different types of current impulse. Furthermore, the damage caused to each sample indicates that, despite masking only the area of interest, the wandering arc on the surface stills plays an important role in distributing the energy input into the CFC and hence the observed damage. Regardless of the different surface damage caused by the different current impulses, the resultant damage from each component current impulse shows polymer degradation with fracturing and lifting up of the carbon fibres. This thesis has then attempted to numerically investigate the physical processes which lead to this lightning strike damage. Within the current state of the art knowledge there is no proposed method to numerically represent the lightning strike arc attachment and the subsequent arc wandering. Therefore, as arc wandering plays an important role in causing the observed damage, it is not possible to numerically model the lightning strike damage. An analogous damage mechanism is therefore needed so the lighting strike damage processes can be numerically investigated. This thesis has demonstrated that damage caused by laser ablation, represents a similar set of physical processes, to those which cause the lightning strike current impulse damage, albeit without any additional electrical processes. Within the numerical model, the CFC is numerically represented through a homogenisation approach and so the relevance and accuracy of a series of analytical methods for predicting the bulk thermal and electrical conductivity for use with CFCs have been investigated. This study has shown that the electrical conductivity is dominated by the percolation effects due to the fibre to fibre contacts. Due to the more comparable thermal conductivity between the polymer and the fibres, the bulk thermal conductivity is accurately predicted by an extension of the Eshelby Method. This extension allows the bulk conductivity of a composite system with more than two composite components to be calculated. Having developed a bespoke thermo-chemical degradation model, a series of validation studies have been conducted. First, the homogenisation approach is validated by numerically investigating the electrical conduction through a two layer panel of CFC. These numerical predictions showed initially unexpected current ow patterns. These predictions have been validated through an experimental study, which in turn validates the application of the homogenisation approach. The novelty within the proposed model is the inclusion of the transport of produced gasses through the decomposing material. The thermo-chemical degradation model predicts that the internal gas pressure inside the decomposing material can reach 3 orders of magnitude greater than that of atmospheric pressure. This explains the de-laminations and fibre cracking observed within the laser ablated damage samples. The numerical predictions show that the inclusion of thermal gas transport has minimal impact on the predicted thermal chemical damage. The numerical predictions have further been validated against the previously obtained laser ablation results. The predicted polymer degradation shows reasonable agreement with the experimentally observed ablation damage. This along with the previous discussions has validated the physical processes implemented within the thermo-chemical degradation model to investigate the thermal chemical lightning strike damage.