The application of new technologies into products allows manufacturers to differentiate their products and fulfil customer requirements. A method to assess the impact of technologies on the fulfilment of customer requirements has been investigated in this research. The main focus is the application of advanced technologies, which may be still in the concept phase, to complex systems using automotive engines as case studies. In these systems the customer will not directly interact with the technologies but the technologies are applied to deliver the range of customer requirements individually or collectively. A standard tool, Quality Functional Deployment (QFD) has been adopted to capture the customer requirements. As all the technology combinations cannot be assessed experimentally due to resource limitations, expert opinion has been employed. A questionnaire has been developed to capture the experts’ opinions of the performance of the technologies in a non-linear Likert-type scale. The combination of expert opinion, where the experts are specialists in a range of engineering disciplines, has been considered and analysis methods incorporating opinion weighting have been developed that parallel established expert opinion analysis methods used for subjective probability risk analysis. The analysis methods enabled different technology combinations to be assessed and customer requirements for different target products to be incorporated to provide a route for optimum technology selection to fulfil customer requirements. A pilot study and two case studies were used to investigate the methodology with Customer feedback used to validate the output in one of the case studies. The research has shown that the Customer Focussed Technology Selection framework and methodology developed in this thesis is a new approach to selecting technologies. The case studies have demonstrated it is an effective framework for evaluating the impact of low technology-readiness-level technologies in complex systems on the Customer. The methodology is suitable for application in technology development processes such as the Jaguar Land Rover TCDS process to support the initial technology selection. / Jaguar Land Rover

Le contact Segment-Piston-Chemise (SPC) est essentiel pour la réduction des émissions de CO2 et de polluants dans les moteurs automobile. Pour optimiser ces deux contraintes antagonistes, il est nécessaire d’étudier l’influence de la microgéométrie afin de trouver de nouvelles architectures de ce contact. L’étude expérimentale étant très coûteuse, il est impératif de pouvoir modéliser les performances du contact SPC. Les méthodes multigrilles permettent de réduire considérablement les temps de calcul. Pour le contact SPC, les méthodes multigrilles classiques perdent leur efficacité en présence de texture. Un nouvel algorithme, basé sur des idées développées par Alcouffe et al. a été conçu qui permet l’étude du contact hydrodynamique en présence de texturation importante. Il a été validé par comparaison avec un modèle analytique 1D, ses performances ont été étudiées et une première étude de l’influence des paramètres sur la portance pour une surface striée a été faite. L’algorithme a ensuite été modifié pour imposer la conservation des flux. Un processus de relaxation plus proche de la physique du phénomène est utilisé. Le code obtenu n’atteint pas l’efficacité multigrille optimale. Cependant sa convergence est suffisamment rapide pour étudier l’influence de pores pour le segment plat et montrer l’importance d’une texturation partielle. Pour finir, le terme transitoire de l’équation de Reynolds est pris en compte. Pour un segment parabolique et une chemise striée, l’influence des paramètres des stries (profondeur, largeur, distance et angle des stries) sur la variation moyenne de la hauteur de film par rapport à une surface lisse est analysée. / The piston ring cylinder liner (PRCL) contact is essential in the CO2 and pollutant emission reduction in internal combustion engine. These two constraints being antagonistic, a compromise has to be found and it is necessary to study the microgeometry influence in order to find new PRCL contact architectures. As experimental study can be very expensive, modeling PRCL contact performance becomes crucial. Multigrid methods allow a huge reduction of the computational time. Unfortunately, in presence of texture, classical multigrid efficiency decays significantly and are not usable. A new algorithm based on an approach developed by Alcouffe et al. was designed that allows the study of the hydrodynamic lubrication of this contact even in the case of a highly varying geometry due to the texture. The obtained code was validated by comparison with a 1D analytical model, its performance was evaluated and a first study of an analytic cross hatched geometry parameter influence on the load carrying capacity was completed. The code was modified to include flow conservation. A relaxation process based closer to the physics phenomenon is used. The code efficiency is not the one that is expected from multigrid technics. However it convergence is sufficiently fast to study the dimple influence for the oil control ring and show the importance of partial texturing. Finally the transient term of the Reynolds equation is accounted for. In the case of a parabolic ring and for a cross hatched liner, the influence of the groove parameters ( depth, width, distance and groove angle) on the average minimum film thickness with respect to the smooth case is analyzed.

[ES] Los objetivos son 2: 1- Determinar el efecto de la dispersión de la recirculación de gases de escape de alta presión (HP EGR) en las emisiones de NOx y humos en motores diésel de automoción en operaciones de funcionamiento constantes. La investigación cuantifica las emisiones de NOx y humos en función del nivel de dispersión de EGR de alta presión entre cilindros. 2- Explorar los límites del modelado 1D para predecir el movimiento del flujo de los gases en la compleja situación en la que estos entran en los cilindros desde el colector de admisión. Los experimentos se realizaron en un banco de pruebas con un motor diésel de 1.6 litros. Para detectar la dispersión de EGR de alta presión se instaló un sistema de válvulas en los conductos de admisión de cada cilindro para medir la concentración de CO2, por tanto la tasa de EGR, en cada conducto. Se instaló también un sistema de válvulas en el escape para medir las emisiones de NOx en cada cilindro. Se instaló un sensor de humos en la línea de escape, aguas abajo de la turbina, para medir el efecto de la dispersión de EGR de alta presión en las emisiones de humos además del sensor para medir el resto de las emisiones contaminantes aguas abajo de la turbina. Se han estudiado 9 puntos de funcionamiento diferentes con distintas velocidades y niveles de carga. El mapa motor se ha estudiado en profundidad, desde 1250 hasta 3000 rpm y entre 3 y 20 bar de presión media efectiva (BMEP). La tasa de EGR varía entre 5 y 42%, dependiendo del punto de funcionamiento. La geometría del modelo reproduce la del motor diésel de automoción de 1.6 litros en el que se realizaron los ensayos experimentales. Incluyendo la línea de EGR de alta presión que fue instalada para controlar los niveles de dispersión durante los ensayos experimentales. La metodología centrada en las herramientas experimentales combina aparatos de medida tradicional con un sistema de válvulas específico que ofrecen una información precisa en cuanto a la concentración de especies tanto en el colector de admisión como en el de escape. El estudio se realizó a emisiones de NOx constantes para observar el efecto de la dispersión de EGR en los valores de opacidad. La metodología está centrada en las herramientas de modelado, las condiciones de contorno y toda la información necesaria para poner en marcha el modelo proviene de los resultados de los ensayos experimentales medidos con los diferentes sensores y aparatos mencionados anteriormente. Muchos de ellos necesarios para ajustar el modelo. La parte más importante para estudiar la capacidad de predicción del modelo es el diseño del colector de admisión. Es necesario poner especial atención en la orientación de los conductos, y en la estructura interna y la superficie para tratar de ser muy fiel a la geometría real, ya que ello determina la predicción de la dispersión. Esta aproximación de modelado cuasi tridimensional (3D) es posible gracias a un programa específico que importa la información necesaria desde un archivo CAD al programa de modelado 1D. Respecto a la parte experimental, el estudio concluye que cuando la dispersión de EGR es baja, los niveles de opacidad se reducen en todos los puntos de funcionamiento. Sin embargo, por encima de ciertos niveles de dispersión de EGR, la opacidad crece seriamente con diferentes pendientes según el punto de operación. El estudio permite cuantificar este límite de dispersión de EGR. La dispersión de EGR incrementa el consumo de combustible por encima del 6.9%. Respecto a la parte de modelado, el estudio concluye que cuando la distribución de EGR entre conductos medida experimentalmente es asimétrica y presenta un alto patrón de concavidad o convexidad, el modelo no predice adecuadamente la distribución del EGR. El estudio concluye que, aunque en los ensayos experimentales la tasa de EGR afecta a la dispersión de EGR, el modelo 1D no es tan sensible como para predecir esta influencia cuando la tasa de EGR está por debajo del 10%. / [CA] L'objectiu de l'estudi és doble. Per una banda, determinar l'efecte de la dispersió de la recirculació de gasos d'escapament d'alta pressió (HP EGR per les seues sigles en anglès) en les emissions d'òxids de nitrogen (NOx) i fums en motors dièsel d'automoció en operacions de funcionament constants. La investigació quantifica les emissions de NOx i fums en funció del nivell de dispersió d'EGR d'alta pressió entre cilindres. Per una altra banda, l'objectiu és explorar els límits del modelatge unidimensional (1D) per predir el moviment del flux dels gasos en la complexa situació en què aquests entren als cilindres des del col·lector d'admissió. Els experiments van ser realitzats en un banc de proves amb un motor dièsel de 1.6 litres. Per detectar la dispersió d'EGR d'alta pressió es va instal·lar un sistema de vàlvules en els conductes d'admissió de cada cilindre per mesurar el percentatge de CO2 i per tant la taxa d'EGR. De la mateixa manera es va instal·lar també un sistema de vàlvules d'escapament, cilindre a cilindre, per mesurar les emissions de NOx. A més també es va instal·lar un sensor de fums en la línia d'escapament, aigües avall de la turbina, per mesurar l'efecte de la dispersió d'EGR d'alta pressió en les emissions de fums, així com el sensor de mesura de la resta d'emissions aigües avall de la turbina. S'han estudiat 9 punts de funcionament diferents amb distintes velocitats i nivells de càrrega, per la qual cosa el mapa motor s'ha estudiat en profunditat, des de 1250 fins a 3000 rpm i entre 3 i 20 bar de pressió mitjana efectiva (BMEP per les seues sigles en anglès). La taxa d'EGR varia entre 5 i 42 %, depenent del punt de funcionament. La geometria del model reprodueix la geometria del motor dièsel d'automoció d'1.6 litres en el qual es van realitzar tots els assajos experimentals. La metodologia centrada en les ferramentes experimentals combina aparells de mesura tradicional amb un sistema de vàlvules específic que ofereixen una informació precisa quant a la concentració d'espècies tant al col·lector d'admissió com al d'escapament. L'estudi es va realitzar a emissions de NOx constants per observar l'efecte de la dispersió d'EGR en els valors d'opacitat. Quant a la metodologia centrada en les ferramentes de modelatge, les condicions de contorn i tota la informació necessària per posar en marxa el model prové dels resultats dels assajos experimentals mesurats amb els diferents sensors i aparells mencionats anteriorment, molts d'ells necessaris per ajustar el model. La part més important per estudiar la capacitat de predicció del model és el disseny del col·lector d'admissió. És necessari posar especial atenció a l'orientació dels conductes, i a l'estructura interna i la superfície per tractar de ser molt fidel a la geometria real, ja que determina la predicció de la dispersió. Esta aproximació del model quasi-tridimensional (3D) és possible gràcies a un programa específic que importa la informació necessària des d'un arxiu de disseny assistit per ordinador (CAD) al programa de modelat 1D. Respecte a la part experimental, l'estudi conclou que quan la dispersió d'EGR és baixa, els nivells d'opacitat es redueixen en tots els punts de funcionament. Tanmateix, per damunt de certs nivells de dispersió d'EGR, l'opacitat creix seriosament amb diferents pendents segons el punt d'operació. L'estudi permet quantificar aquest límit de dispersió d'EGR. A més, la dispersió d'EGR podria contribuir a incrementar el consum de combustible per damunt del 6.9%. Respecte a la part de modelatge, l'estudi conclou que quan la distribució d'EGR entre conductes mesurada experimentalment és asimètrica i presenta un alt patró de concavitat o convexitat, el model no prediu adequadament la distribució d'EGR. A més, l'estudi conclou que, tot i que en els assajos experimentals la taxa d'EGR afecta a la dispersió d'EGR, el model 1D no és tan sensible com per predir aquesta influència quan la taxa d’EGR està per baix del 10%. / [EN] The objective of the study is twofold. On the one hand, it is to determine the effect of the high pressure (HP) exhaust gas recirculation (EGR) dispersion in automotive diesel engines on NOx and smoke emissions in steady engine operation. The investigation quantifies the smoke emissions as a function of the dispersion of the HP EGR among cylinders. On the other hand, it is to explore the limits of the one-dimensional (1D) modeling to predict the movement of the flow in a complex situation as the gases get into the cylinders from the intake manifold. The experiments are performed on a test bench with a 1.6 liter automotive diesel engine. In order to track the HP EGR dispersion in the intake pipes, a valves system to measure CO2, hence EGR rate, pipe to pipe was installed. In the same way, a valves device to measure NOx emissions cylinder to cylinder in the exhaust was installed too. Moreover a smoke meter device was installed in the exhaust line, downstream the turbine, to measure the effect of the HP EGR dispersion on smoke emissions. A probe to measure the other raw emissions was installed downstream the turbine, too. Nine different engine running conditions were studied at different speed and load, thus the engine map was widely studied, from 1250 rpm to 3000 rpm and between 3 and 20 bar of BMEP. The EGR rate variates between 5 and 42 % depending on the working operation point. The geometry of the model reproduces the geometry of a 1.6 liter diesel automotive engine where the tests were performed. It includes an HP-EGR line and the device that was installed to perform the experiments to control the dispersion. The methodology focused on experimental tools combining traditional measuring devices with a specific valves system which offers accurate information about species concentration in both the intake and the exhaust manifolds. The study was performed at constant raw NOx emissions to observe the effect of the EGR dispersion in the opacity values. Regarding the methodology focused on modeling tools, the boundary conditions and all the necessary information to run the model comes from experimental results measured with the different sensors and devices mentioned before. Much of them were needed to adjust the model. The most important part of the modeling to study the capacity of the prediction of the EGR dispersion is the layout of the intake manifold. It is necessary put special attention to the orientation of the pipes, and the internal structure and surface trying to mimic the real geometry because it determinates the prediction of the dispersion. This approximation to quasi-three-dimensional (3D) modeling is possible thanks to a specific software that imports the necessary information from a computer-aided design (CAD) file to the 1D modeling software. Concerning the experimental results, the study leads to conclude that when the EGR dispersion is low, the opacity presents reduced values in all operation points. However, above a certain level of EGR dispersion, the opacity increases dramatically with different slopes depending on the engine running condition. This study allows quantifying this EGR dispersion threshold. In addition, the EGR dispersion could contribute to an increase in the engine fuel consumption up to 6.9%. Regarding to the modeling part, the study concludes that when the experimental EGR distribution among pipes is asymmetric and presents high concavity or convexity spatial pattern, the model does not predict properly the EGR distribution. In addition, the study concludes that, although in the experimental tests the EGR rate affects to the EGR dispersion, the 1D model is not too sensitive to predict this influence when the EGR rate is lower than 10%. / The respondent wishes to acknowledge the financial support received by contract FPI 2015 S2 3101 of Programa de Apoyo a la Investigación y Desarrollo (PAID) from Universitat Politècnica de València (UPV). / Miguel García, J. (2021). Analysis of the high pressure EGR dispersion among cylinders in automotive diesel engines [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/161889 / TESIS

The compression behavior of a newly developed near eutectic Al-Si based cast alloy with three different microstructures has been investigated in the present work. Microstructures with modified and unmodified Si particles and matrix with different tempers are investigated. The main objective of this work is to understand the effect of heat treatment and modification on the fracture behavior of the alloy under compression. This alloy is subjected to compressive loading at different strain rates and temperatures during the operation of the engines. Hence, the effect of strain rates and temperatures is also considered. The compression tests are carried out at different strain rates from quasi-static to dynamic viz., 3*10-4 to 102/s and three different temperatures RT, 100°C and 200°C. Microstructure of the alloys studied predominantly consists of eutectic colonies of α-Al and Si with a few interspersed α-Al dendrites. Modified alloy has more globular Si particles than unmodified alloy. Heat treated alloys are found to have hardening precipitates S’ & Al7Cu4Ni and 3-7 atomic layer thick zones, which may be precursors to S’ phase. A variety of large intermetallics, viz., plate like particles Al4.5FeSi, Chinese script like particles Al19Fe4MnSi2 and bulky phase Al3NiCu are also observed in the alloys. Mechanical behavior of the alloys is found to be different for different microstructures. Modification improves strength and ductility. Heat treatment improves strength of the alloy at the expense of ductility. A transition in mechanical behavior is observed after a particular strain rate for all the alloys studied. This transition strain rate is dependent on heat treatment, Si particle size and temperature. This transition can be explained on the basis of dislocation-precipitate and dislocation-Si particle interactions. Work hardening behavior of the alloys depends on the matrix microstructure in the unmodified alloys, and both matrix and particles play a role in the modified alloy. A statistically robust quantitative micro structural analysis has been carried out after compressing the samples at various strain rates and temperatures. The unique contribution of this work is the understanding of combined effect of strain rate and temperature on Si particle fracture characteristics in the alloy with different microstructures. From the fracture characteristics of Si particles, it is concluded that both dislocation pile-up mechanism and fibre loading are responsible for particle fracture in the modified alloy, whereas the fibre loading mechanism alone is sufficient to explain the particle fracture characteristics in the unmodified alloy. Si particles in the modified condition are found to cleave along the lowest surface energy planes {112} & {110} and the particles with orientations {112} & {111} are more prone to fracture. In addition to Si particle fracture, elongated Fe rich intermetallic particles are also seen to show peculiar fracture behavior. The Al4.5FeSi intermetallics with (100) as the plane of the plate cleave along (100) planes. This is a novel finding in this work and could have immense implications on the role of Fe impurities in the fracture behavior of these alloys. Moreover, since these cleavage fractures are seen to be more than 200 microns in size (which implies that the real penny shaped crack would be even larger) their role cannot be assumed to be negligible, as was previously thought. The load sharing between the Al matrix and eutectic Si particles are simulated by microstructure based finite element modeling. The program OOF (Object-Oriented Finite element analysis) is used to generate the finite element meshes for real microstructures with different Si morphology. The experimentally obtained stress – strain properties of the alloy is given as an input to describe the plastic behavior of the Al matrix, in the finite element simulation. This analysis helps to understand the effect of particle size, shape, orientation & clustering and matrix temper on the stress transferred to the Si particles. Combination of Electron Back-Scattered Diffraction (EBSD) and frequency shift, polarized micro-Raman technique is applied to validate the stress states in Si particles with {111} orientations. The stress at fracture of Si particles is also estimated from Raman technique. Even though the alloys with different microstructures show different mechanical behavior, the sequence of fracture mechanisms is found to be same for all the alloys. The failure occurs in three stages: cracking of Si particles at low strains, micro-crack formation along the fractured particles, micro-crack coalescence and propagation leading to final failure. Thus, the proposed analysis links various deformation mechanisms ranging from nano precipitate-dislocation interactions to micro short-fiber theory of load sharing by eutectic silicon along with coupled effect of strain rate and temperature. In addition, negative strain rate sensitivity is also observed in the lower strain rate regimes (3*10-4, 10-3& 102/s) at RT and 100°C for all the three alloys, and serrated flow is also observed in the same strain rate and temperature regimes. Some of the features of serrated flow can be explained by the dynamic strain aging model and some other features by precipitate shearing.