Modeling of electrical machines play a very important role in a variety of applications such as performance analysis, characterization, fault diagnosis, condition monitoring and stress analysis of the machines. One of the important parameter while modeling the electrical machine is the magnetic characteristics of the core material. This plays a huge role in the performance characteristics and analysis of the electrical machines. Existing techniques available to determine the magnetic characteristics of a material are mainly Epstein and single sheet tester. These two tests are invasive and destructive method of testing the magnetic characteristics of the material. This research work takes up this problem and comes with a simple yet effective solution to determine the average magnetic characteristics of the material in the salient pole synchronous machine (SPSM). An FE model of the SPSM was developed to closely emulate the characteristics of the experimental machine. This FE model was first subjected to magnetostatic simulation under different field currents using a known magnetic material. Ampere turn distribution technique was used to determine the magnetic characteristics of the material. Following the determination of the new material, this material was used in the FE simulation of the SPSM running as a motor and a generator under varying load condition and field currents. Then these results were compared with the real machine to determine the effectiveness of the developed scheme. / Graduate

Hip fracture has long been considered as the most serious consequence of osteoporosis, which includes chronic pain, disability, and even death. In the elderly population, a femur fracture is very common. It is assessed that 50% of women aged 50 or older may experience a hip fracture in their remaining life. Hip fracture is among the most common injuries and can lead to substantial morbidity and mortality. In the US alone, over 250,000 hip fractures occur each year and this number is expected to double by the year 2040. Statistics indicate that over 20% of people who experience a hip fracture die within one year and only 25% have a total recovery. Femur fractures are now becoming a major social and economic burden on the health care system. In practice, it is very difficult to predict the femur fracture risks. One of the main reasons is that there is not a robust and easy-to-get measure to quantify the strength of the bone. Clinicians use bone mineral density (BMD) as an indicator of osteoporosis and fracture risk. Several studies showed that BMD cannot be used alone to identify bone strength. In fact, the majority of patients who suffer from fractures have normal or even higher BMD scores. There are a large number of risk factors that contribute to the occurrence of femur fracture, which should also be involved in predicting hip fracture risks. For example, age, weight, height, ethnicity and so on. Some of the factors might not have been identified yet. Thus, there will be a high level of uncertainty in the clinical dataset, which makes it difficult to construct and validate a hip risk prediction model. The objective of the dissertation is to construct an improved hip fracture risk prediction model. Due to the difficulty of obtaining experimental or clinical data, computational simulations might help increase the predictive ability of the risk model. In this research, the hip fracture risk model is based on a support vector machine (SVM) trained using a clinical dataset from the Women's Health Initiative (WHI). In order to improve the SVM-based hip fracture risk model, data from a fully parameterized finite element (FE) model is used to supplement the clinical dataset. This FE model allows one to simulate a wide range of geometries and material properties in the hip region, and provides a measure of risk based on mechanical quantities (e.g., strain). This dissertation presents new approaches to fuse the clinical data with the FE data in order to improve the predictive capability of the hip fracture risk prediction model. Two approaches are introduced in this dissertation to construct a hybrid risk model: an "augmented space" approach and a "computational patients" approach. This work has led to the construction of a new online hip fracture risk calculator with free access.

<p>Explosions are associated with more than 80% of the casualties in the Iraq and Afghanistan wars. Given the widespread use of thoracic protective armor, the most prevalent injury for combat personnel is blast-related traumatic brain injury (TBI). Almost 20% of veterans returning from duty had one or more clinically confirmed cases of TBI. In the decades of research prior to 2000, neurotrauma was under-recognized as a blast injury and the etiology and pathology of these injuries remains unclear.</p><p>This dissertation used the finite element (FE) method to address many of the biomechanics-based questions related to blast brain injuries. FE modeling is a powerful tool for studying the biomechanical response of a human or animal body to blast loading, particularly because of the many challenges related to experimental work in this field. In this dissertation, novel FE models of the human and ferret head were developed for blast and blunt impact simulation, and the ensuing response of the brain was investigated. The blast conditions simulated in this research were representative of peak overpressures and durations of real-world explosives. In general, intracranial pressures were dependent on the peak pressure of the impinging blast wave, but deviatoric responses in the brain were dependent on both peak pressure and duration. The biomechanical response of the ferret brain model was correlated with in vivo injury data from shock tube experiments. This accomplishment was the first of its kind in the blast neurotrauma field.</p><p>This dissertation made major contributions to the field of blast brain injury and to the understanding of blast neurotrauma. This research determined that blast brain injuries were brain size-dependent. For example, mouse-sized brains were predicted to have approximately 7 times larger brain tissue strains than the human-sized brains for the same blast exposure. This finding has important implications for in vivo injury model design, and a scaling model was developed to relate animal experimental models to humans via scaling blast duration by the fourth-root of the ratio of brain masses. </p><p>This research also determined that blast neurotrauma is correlated to deviatoric metrics of the brain tissue rather than dilatational metrics. In addition, strains in the blasted brain were an order-of-magnitude lower than expected to produce injury with traditional closed-head TBI, but an order-of-magnitude higher in strain rate. The 50th percentile peak principle strain metric of values of 0.6%, 1.8%, and 1.6% corresponded to the 50% risk of mild brain bleeding, moderate brain bleeding, and apnea respectively. These findings imply that the mechanical thresholds for brain tissue are strain-based for primary blast injury, and different from the thresholds associated with blunt impact or concussive brain injury because of strain rate effects.</p><p>The conclusions in this dissertation provide an important guide to the biomechanics community for studying neurotrauma using in vivo, in vitro, and in silico models. Additionally, the injury risk curves developed in this dissertation provide an injury risk metric for improving the effectiveness of personal protective equipment or evaluating neurotrauma from blast.</p> / Dissertation

Mitigating injury in side impact has been an important topic of research for decades. In the mid 1980’s the American government began a program intended to improve the crashworthiness of vehicles in side impact. This program ultimately led to the introduction of a dynamic side impact test (Federal Motor Vehicle Safety Standard (FMVSS) 214), which new vehicles must pass, along with a very similar test aimed at consumer awareness (New Car Assessment Program (NCAP) side impact test). The work presented in this thesis involved the study and simulation of these tests to evaluate occupant response in side impact, with a focus on the thoracic response. In the first portion of the work presented here, an in-depth study of the National Highway Traffic Safety Administration (NHTSA) crash test database was performed. In this study the results of the side impact crash tests of 72 vehicles were examined to understand the general trends seen in this type of testing with regards to vehicle velocity, side intrusion, and occupant injury prediction. A series of average velocity profile curves was created from accelerometer data at 18 measurement points on each vehicle crash tested. Additionally the injury criterion measured by the front seat occupant was plotted against several vehicle variables (such as mass and occupant arm to door distance) to study the effect these variable had on the injury predicted by the occupant. No single variable was shown to have a strong correlation to injury, although increasing door intrusion distance, peak lateral velocity, the Head Injury Criterion (HIC), and pelvic acceleration were found to positively correlate to thoracic injury. In addition, increasing vehicle model year, vehicle mass, and arm to door (AD) distance showed negative correlations with thoracic injury. Following the survey of the NHTSA database, a finite element model of the NHTSA side impact test was developed. This model included a full scale Ford Taurus model, a NHTSA barrier model and three side impact anthropometric test device (ATD) occupant models, each representing a different 50th percentile male dummy. Validation of this model was carried out by comparing the simulated vehicle component velocity results to the corridors developed in the NHSTA crash test database study as well as comparing these velocities, the vehicle deformation profile, and the occupant velocity, acceleration and rib deflection to several Ford Taurus crash tests from a similar vintage to the finite element model. As this model was intended as a ‘baseline’ case to study side impact and occupant kinematics in side impact, side airbags were not included in this model. A lack of experimental data and a lack of consensuses within the automotive crash community on the proper method of modeling these devices and their effectiveness in real world impacts also led to their exclusion. Following model validation, a parametric study was carried out to assess the importance of the initial position of the occupant on the vehicle door velocity profile and the predicted occupant injury response. Additionally the effect of the door trim material properties, arm rest properties and the effect of seat belt use were studied. It was found that the lateral position of the occupant had an effect on the door velocity profile, while the vertical and longitudinal position did not. The use of seatbelts was shown to have no significant effect in these simulations, due to minimal interaction between the restraint system and occupant during side impact. Furthermore, there was a general decreasing trend in the injury predicted as the initial position of the occupant was moved further inboard, down and forward in the vehicle. Stiffer interior trim was found to improve the injury prediction of the occupant, while changing the material of the foam door inserts had no effect. It was found that in general the occupant remained in position, due to the inertia of the occupant, while the seat began moving towards the centerline of the vehicle. Future considerations could include more advanced restraint systems to couple the occupant more effectively to the seat, or to develop side interior trim that engages the occupant earlier to reduce the relative velocity between the occupant and intruding door. Overall, the model correlated well with experimental data and provided insight into several areas which could lead to improved occupant protection in side impact. Future work should include integrating side airbags into the model, widening the focus of the areas of injury to include other body regions and integrating more detailed human body models.

Shape memory alloys and polymers have been extensively researched recently because of their unique ability to recover large deformations. Shape memory polymers (SMPs) are able to recover large deformations compared to shape memory alloys (SMAs), although SMAs have higher strength and are able to generate more stress during recovery. This project focuses on procedure for fabrication and Finite Element Modeling (FEM) of a shape memory composite actuator. First, SMP was characterized to reveal its mechanical properties. Specifically, glass transition temperature, the effects of temperature and strain rate on compressive response and recovery properties of shape memory polymer were studied. Then, shape memory properties of a NiTi wire, including transformation temperatures and stress generation, were investigated. SMC actuator was fabricated by using epoxy based SMP and NiTi SMA wire. Experimental tests confirmed the reversible behavior of fabricated shape memory composites. The Finite Element Method was used to model the shape memory composite by using a pre-written subroutine for SMA and defining the linear elastic and plastic properties of SMP. ABQUS software was used to simulate shape memory behavior. Beside the animated model in ABAQUS, constitutive models for SMA and SMP were also developed in MATLAB® by using the material properties obtained from experiments. The results of FEM simulation of SMC were found to be in good agreement with experimental results.

Heart failure is an important medical disease and impacts millions of people throughout the world. In order to treat this problem, biomaterial injectable treatment injected into the myocardium of the failing LV are currently being developed. Through this treatment, the biomaterial material injections can reduce wall stresses during the cardiac remodeling process. By using computational techniques to analyze the effects of a treatment involving the injection of biomaterial material into the LV after MI, the material parameters of the hydrogel injections can be optimized. The results shows that the hydrogel injections could reduce the global average fiber stress and the transmural average stress seen from optimization. These results indicated that the hydrogel injections could influence the stiffness in passive LV tissue, but there is still need for more research on the active part of ventricular contraction. Conclusion: hydrogel injection is a viable way to alter ventricular mechanical properties.

Myocardial infarctions induce a maladaptive ventricular remodeling process that independently contributes to heart failure. In order to develop effective treatments, it is necessary to understand the way and extent to which the heart undergoes remodeling over the course of healing. There have been few studies to produce any data on the in-vivo material properties of infarcts, and much less on the properties over the time course of healing. In this paper, the in-vivo passive material properties of an infarcted porcine model were estimated through a combined use of magnetic resonance imaging, catheterization, finite element modeling, and a genetic algorithm optimization scheme. The collagen fiber orientation at the epicardial and endocardial surfaces of the infarct were included in the optimization. Data from porcine hearts (N=6) were taken at various time points after infarction, specifically 1 week, 4 weeks, and 8 weeks post-MI. The optimized results shared similarities with previous studies. In particular, the infarcted region was shown to dramatically increase in stiffness at 1 week post-MI. There was also evidence of a subsequent softening of the infarcted region at later time points post infarction. Fiber orientation results varied greatly but showed a shift toward a more circumferential orientation.

Additive manufacturing (AM) offers reduced material waste and energy usage, as well as an increase in precision. Direct metal AM is used not only for prototyping, but to produce final production parts in the aerospace, medical, automotive and other industries. Process mapping is an approach that represents process outcomes in terms of process input variables. Solidification microstructure process maps are developed for single bead and thin wall deposits of Ti-6Al-4V via an electron beam wire feed and electron beam powder bed AM process. Process variable combinations yielding constant beta grain size and morphology are identified. Comparison with the process maps for melt pool geometry shows that by maintaining a constant melt pool cross sectional area, a constant grain size will also be achieved. Additionally, the grain morphology boundaries are similar to curves of constant melt pool aspect ratio. Experimental results are presented to support the numerical predictions and identify a proportional size scaling between beta grain widths and melt pool widths. Results demonstrate that in situ, indirect control of solidification microstructure is possible through direct melt pool dimension control. The ability to control solidification microstructure can greatly accelerate AM process qualification potentially allow for tailored microstructure to the desired application.

Pour optimiser les structures des aéronefs, il est maintenant nécessaire de concevoir le matériau au « juste-besoin », de façon à diminuer le ratio masse/performances. Par une bonne gestion du procédé de fabrication et un choix judicieux des matériaux constitutifs, les composites à renfort tissé et à matrice organique ont ce potentiel. Mais pour l’exploiter pleinement, de nouvelles approches adaptées à ce type de matériau doivent être développées. Pour cela, une chaîne de calcul multimodèle est proposée, permettant de prévoir les propriétés mécaniques élastiques saines ou endommagées du matériau à partir de ses paramètres de conception. Cette chaîne est établie à l’échelle mésoscopique, pour pouvoir prendre en compte la géométrie du renfort. Une procédure spéciale de création de maillages de cellules mésoscopiques de composites tissés a été développée, de façon à faire le lien entre la déformée du renfort après mise en forme, obtenue par simulation EF, et les autres modèles de la chaîne (injection de résine, cuisson du composite, comportement mécanique). Le bon fonctionnement de l’approche est montré par l’étude de deux cas-tests, un renfort de quatre plis de taffetas et un renfort de quatre plis de satin de 5, chacun compactés à différents niveaux et selon plusieurs configurations d’imbrication de plis. Enfin, pour anticiper la validation de la chaîne de modélisation, une étude expérimentale comparative entre plusieurs composites tissés compactés à différentes épaisseurs a été menée. Ce travail se place dans le cadre de la construction future d’une chaîne multiéchelle plus globale qui, parcourue dans le sens inverse, permettra de concevoir le matériau sur-mesure en fonction des performances structurales locales désirées. / In order to optimize aeronautic structures, the manufacturing process must be tailored to the structural needs, with the aim of reducing the density/performance ratio. Polymer composites with woven reinforcements offer a large flexibility due to a vast choice of constituent materials and manufacturing process parameters. However, to entirely exploit their potential, new design methods specifically adapted to this type of material have to be developed. For this purpose, a modeling chain is proposed, which is able to predict the elastic properties of the intact or damaged material, by incorporating the manufacturing process parameters. The chain is built at the mesoscopic scale, to take into account the reinforcement geometry. A special procedure to generate finite element (FE) meshes of mesoscopic representative unit cells of woven composites has been developed, which links the deformation of the reinforcement, obtained from FE calculations, to the other models of the chain (resin injection, curing, and mechanical behavior). Two materials are studied to show the potential of the modeling chain: A four ply lay-up of a plain weave and of a satin weave fabric are considered, each of them having several compaction ratios and different nesting between the plies. With the aim of a validation of the modeling chain, multi-instrumented experimental tests have been carried out on several multi-layer plain weave composites with different thicknesses. In future applications, the proposed strategy will be placed in a toolbox able to design optimum woven composite structures based on local performance requirements.

A análise da formação e crescimento de fissuras em peças de concreto armado permanece como uma das principais dificuldades no campo da engenharia estrutural. Considerando que as fissuras têm uma influência muito grande no comportamento estrutural global, estudos para prever e controlar a fissuração do concreto são de essencial importância. O objetivo deste trabalho é apresentar um modelo numérico do tipo incorporado para representar as fissuras em peças de concreto armado submetidas aos esforços de flexão e corte, ou seja, um modelo que seja capaz de simular, além das fissuras perpendiculares ao eixo da peça, fissuras inclinadas. Os modelos de fissura incorporada se baseiam no conceito de descontinuidades incorporadas dentro de elementos finitos padrões. No modelo empregado neste trabalho, a fissura é representada através de uma descontinuidade no campo interno de deslocamentos do elemento. O modelo incorporado implementado é uma continuação do trabalho desenvolvido por d’Avila, que baseou-se no modelo de Dvorkin Cuitiño e Gioia que, por sua vez, não inclui a contribuição da armadura no equilíbrio interno de forças do elemento. A interação entre as barras de aço e o concreto é simulada através um modelo de transferência de tensão por aderência entre os dois materiais, conforme Russo, Zingone e Romano e FIB - Bulletin 10. Para representar o comportamento do concreto intacto, utiliza-se o modelo constitutivo de Ottosen. Já para representar as barras de aço da armadura, emprega-se o modelo incorporado desenvolvido por Elwi e Hrudey, que permite uma disposição arbitrária das barras de aço no interior dos elementos de concreto. O modelo constitutivo adotado para a armadura é do tipo elasto-plástico com endurecimento. Foi possível simular a fissuração em flexão e corte em vigas de concreto armado com boa correlação com resultados experimentais. Tais situações não poderiam ser analisadas pelo modelo básico sem as modificações propostas nesta dissertação. / The analysis of the formation and growth of cracks in reinforced concrete members remains as one of the main difficulties in the field of structural engineering. Considering that the crack has a considerable influence in the global structural behavior, studies to predict and to control concrete cracking are of essential importance. The aim of this work is to present a numerical model of the embedded type to represent the cracks in reinforced concrete members under bending and shearing efforts, i. e. , a model that is capable to simulate not only cracks that are perpendicular to the axle of the members but also inclined cracks. The embedded crack models are based on the concept of incorporated discontinuities inside of standard finite elements. In the model used in this work, the crack is represented by a discontinuity in the internal field of the element displacements. The embedded model proposed is a continuation of the work developed by d’Avila, which is based on the model of Dvorkin, Cuitiño e Gioia, that does not consider the inclusion of the reinforced contribution in the internal force equilibrium of the element. A bond stress-transfer approach is used to include this reinforcement contribution. To represent the behavior of the uncracked concrete, the Ottosen constitutive model was used. The embedded model presented by Elwi and Hrudey was employed to represent the reinforcement bars, that allows an arbitrary disposal of the bars of steel inside of the concrete elements. The constitutive model adopted for reinforcement is elasto-plastic with hardening. It was possible to simulate the cracking in bending and shearing in reinforced concrete beams with good agreement with experimental results. These cases could not be analyzed by the basi model without the present proposed modifications.