Two-stroke poppet-valved engines may combine the high power density of two - stroke engines and the low emissions of poppet-valved engines. A two-stroke diesel engine can generate the same power as a four-stroke engine of the same size, but at higher (leaner) air/fuel ratios. Diesel combustion at high air/fuel ratios generally means hydrocarbons, soot and carbon monoxide are oxidised more completely to water and carbon dioxide in the cylinder, and the opportunity to increase the rate of exhaust gas recirculation should reduce the formation of nitrogen oxides (NOx). The concept is being explored as a means of economically modifying diesel engines to make them cleaner and/or more powerful. This study details the application of two computational models to this problem. The first model is a relatively simple thermodynamic model created by the author capable of rapidly estimating the behaviour of entire engine systems. It was used to estimate near-optimum engine system parameters at single engine operating points and over a six-mode engine cycle. The second model is a detailed CFD model called KIVA-ERC. It is a hybrid of the KIVA engine modelling package developed at the Los Alamos National Laboratory and combustion and emissions subroutines developed at the University of Wisconsin-Madison Engine Research Center. It was used for detailed scavenging and combustion simulations and to provide estimates of emissions levels. Both models were calibrated and validated for four-stroke cycle operation using experimental data. The thermodynamic model was used to provide initial and boundary conditions to the KIVA-ERC model. Conversely, the combustion simulations were used to adjust zero-dimensional combustion correlations when experimental data was not available. Scavenging simulations were performed with shrouded and unshrouded intake valves. A new two-zone scavenging model was proposed and validated using multidimensional scavenging simulations. A method for predicting the behaviour of the two-stroke engine system based on four-stroke data has been proposed. The results using this method indicate that a four-stroke diesel engine with minor modifications can be converted to a two-stroke cycle and achieve substantially the same fuel efficiency as the original engine. However, emissions levels can not be predicted accurately without experimental data from a physical prototype. It is therefore recommended that such a prototype be constructed, based on design parameters obtained from the numerical models used in this study.

The data-driven models used for the design of powertrain controllers are typically based on the data obtained from steady-state experiments. However, they are only valid under stable conditions and do not provide any information on the dynamic behaviour of the system. In order to capture this behaviour, dynamic modelling techniques are intensively studied to generate alternative solutions for engine mapping and calibration problem, aiming to address the need to increase productivity (reduce development time) and to develop better models for the actual behaviour of the engine under real-world conditions. In this thesis, a dynamic modelling approach is presented undertaken for the prediction of NOx emissions for a 2.0 litre Diesel engine, based on a coupled pre-validated virtual Diesel engine model (GT- Suite ® 1-D air path model) and in-cylinder combustion model (CMCL ® Stochastic Reactor Model Engine Suite). In the context of the considered Engine Simulation Framework, GT Suite + Stochastic Reactor Model (SRM), one fundamental problem is to establish a real time stochastic simulation capability. This problem can be addressed by replacing the slow combustion chemistry solver (SRM) with an appropriate NOx surrogate model. The approach taken in this research for the development of this surrogate model was based on a combination of design of dynamic experiments run on the virtual diesel engine model (GT- Suite), with a dynamic model fitted for the parameters required as input to the SRM, with a zonal design of experiments (DoEs), using Optimal Latin Hypercubes (OLH), run on the SRM model. A response surface model was fitted on the predicted NOx from the SRM OLH DoE data. This surrogate NOx model was then used to replace the computationally expensive SRM simulation, enabling real-time simulations of transient drive cycles to be executed. The performance of the approach was validated on a simulated NEDC drive cycle, against experimental data collected for the engine case study. The capability of methodology to capture the transient trends of the system shows promising results and will be used for the development of global surrogate prediction models for engine-out emissions.

The full text will be available at the end of the embargo period: 4th Aug 2025

In this thesis, exergy based SI engine model optimisation (EBSIEMO) is studied and evaluated. A four-stroke bi-fuel spark ignition (SI) engine is modelled for optimisation of engine performance based upon exergy analysis. An artificial neural network (ANN) is used as an emulator to speed up the optimisation processes. Constrained particle swarm optimisation (CPSO) is employed to identify parameters such as equivalence ratio and ignition time for optimising of the engine performance, based upon maximising 'total availability'. In the optimisation process, the engine exhaust gases standard emission were applied including brake specific CO (BSCO) and brake specific NOx (BSNOx) as the constraints. The engine model is developed in a two-zone model, while considering the chemical synthesis of fuel, including 10 chemical species. A computer code is developed in MATLAB software to solve the equations for the prediction of temperature and pressure of the mixture in each stage (compression stroke, combustion process and expansion stroke). In addition, Intake and exhaust processes are calculated using an approximation method. This model has the ability to simulate turbulent combustion and compared to computational fluid dynamic (CFD) models it is computationally faster and efficient. The selective outputs are cylinder temperature and pressure, heat transfer, brake work, brake thermal and volumetric efficiency, brake torque, brake power (BP), brake specific fuel consumption (BSFC), brake mean effective pressure (BMEP), concentration of CO2, brake specific CO (BSCO) and brake specific NOx (BSNOx). In this model, the effect of engine speed, equivalence ratio and ignition time on performance parameters using gasoline and CNG fuels are analysed. In addition, the model is validated by experimental data using the results obtained from bi-fuel engine tests. Therefore, this engine model was capable to predict, analyse and useful for optimisation of the engine performance parameters. The exergy based four-stroke bi-fuel (CNG and gasoline) spark ignition (SI) engine model (EBSIEM) here is used for analysis of bi-fuel SI engines. Since, the first law of thermodynamic (the FLT), alone is not able to afford an appropriate comprehension into engine operations. Therefore, this thesis concentrates on the SI engine operation investigation using the developed engine model by the second law of thermodynamic (the SLT) or exergy analysis outlook (exergy based SI engine model (EBSIEM)) In this thesis, an efficient approach is presented for the prediction of total availability, brake specific CO (BSCO), brake specific NOx (BSNOx) and brake torque for bi-fuel engine (CNG and gasoline) using an artificial neural network (ANN) model based on exergy based SI engine (EBSIEM) (ANN-EBSIEM) as an emulator to speed up the optimisation processes. In the other words, the use of a well trained an ANN is ordinarily much faster than mathematical models or conventional simulation programs for prediction. The constrained particle swarm optimisation (CPSO)-EBSIEM (EBSIEMO) was capable of optimising the model parameters for the engine performance. The optimisation results based upon availability analysis (the SLT) due to analysing availability terms, specifically availability destruction (that measured engine irreversibilties) are more regarded with higher priority compared to the FLT analysis. In this thesis, exergy based SI engine model optimisation (EBSIEMO) is studied and evaluated. A four-stroke bi-fuel spark ignition (SI) engine is modelled for optimisation of engine performance based upon exergy analysis. An artificial neural network (ANN) is used as an emulator to speed up the optimisation processes. Constrained particle swarm optimisation (CPSO) is employed to identify parameters such as equivalence ratio and ignition time for optimising of the engine performance, based upon maximising 'total availability'. In the optimisation process, the engine exhaust gases standard emission were applied including brake specific CO (BSCO) and brake specific NOx (BSNOx) as the constraints. The engine model is developed in a two-zone model, while considering the chemical synthesis of fuel, including 10 chemical species. A computer code is developed in MATLAB software to solve the equations for the prediction of temperature and pressure of the mixture in each stage (compression stroke, combustion process and expansion stroke). In addition, Intake and exhaust processes are calculated using an approximation method. This model has the ability to simulate turbulent combustion and compared to computational fluid dynamic (CFD) models it is computationally faster and efficient. The selective outputs are cylinder temperature and pressure, heat transfer, brake work, brake thermal and volumetric efficiency, brake torque, brake power (BP), brake specific fuel consumption (BSFC), brake mean effective pressure (BMEP), concentration of CO2, brake specific CO (BSCO) and brake specific NOx (BSNOx). In this model, the effect of engine speed, equivalence ratio and ignition time on performance parameters using gasoline and CNG fuels are analysed. In addition, the model is validated by experimental data using the results obtained from bi-fuel engine tests. Therefore, this engine model was capable to predict, analyse and useful for optimisation of the engine performance parameters. The exergy based four-stroke bi-fuel (CNG and gasoline) spark ignition (SI) engine model (EBSIEM) here is used for analysis of bi-fuel SI engines. Since, the first law of thermodynamic (the FLT), alone is not able to afford an appropriate comprehension into engine operations. Therefore, this thesis concentrates on the SI engine operation investigation using the developed engine model by the second law of thermodynamic (the SLT) or exergy analysis outlook (exergy based SI engine model (EBSIEM)) In this thesis, an efficient approach is presented for the prediction of total availability, brake specific CO (BSCO), brake specific NOx (BSNOx) and brake torque for bi-fuel engine (CNG and gasoline) using an artificial neural network (ANN) model based on exergy based SI engine (EBSIEM) (ANN-EBSIEM) as an emulator to speed up the optimisation processes. In the other words, the use of a well trained an ANN is ordinarily much faster than mathematical models or conventional simulation programs for prediction. The constrained particle swarm optimisation (CPSO)-EBSIEM (EBSIEMO) was capable of optimising the model parameters for the engine performance. The optimisation results based upon availability analysis (the SLT) due to analysing availability terms, specifically availability destruction (that measured engine irreversibilties) are more regarded with higher priority compared to the FLT analysis.

At the end of the 19th Century, the invention of the Internal Combustion Engine (ICE) marked the beginning of our current lifestyle. Soon after the first ICE patent, the importance of increasing air pressure upstream the engine cylinders was revealed. At the beginning of the 20th Century turbo-machinery developments (which had started time before), met the ICE what represented the beginning of turbocharged engines. Since that time, the working principle has not fundamentally changed. Nevertheless, stringent emissions standards and oil depletion have motivated engine developments; among them, turbocharging coupled with downsized engines has emerged as the most feasible way to increase specific power while reducing fuel consumption. Turbocharging has been traditionally a complex problem due to the high rotational speeds, high temperature differences between working fluids (exhaust gases, compressed air, lubricating oil and cooling liquid) and pulsating flow conditions. To improve current computational models, a new procedure for turbochargers characterization and modelling has been presented in this Thesis. That model divides turbocharger modelling complex problem into several sub-models for each of the nonrecurring phenomenon; i.e. heat transfer phenomena, friction losses and acoustic non-linear models for compressor and turbine. A series of ad-hoc experiments have been designed to aid identifying and isolating each phenomenon from the others. Each chapter of this Thesis has been dedicated to analyse that complex problem proposing different sub-models. First of all, an exhaustive literature review of the existing turbocharger models has been performed. Then a turbocharger 1-D internal Heat Transfer Model (HTM) has been developed. Later geometrical models for compressor and turbine have been proposed to account for acoustic effects. A physically based methodology to extrapolate turbine performance maps has been developed too. That model improves turbocharged engine prediction since turbine instantaneous behaviour moves far from the narrow operative range provided in manufacturer maps. Once each separated model has been developed and validated, a series of tests considering all phenomena combined have been performed. Those tests have been designed to check model accuracy under likely operative conditions. The main contributions of this Thesis are the development of a 1-D heat transfer model to account for internal heat fluxes of automotive turbochargers; the development of a physically-based turbine extrapolation methodology; the several tests campaigns that have been necessary to study each phenomenon isolated from others and the integration of experiments and models in a comprehensive characterization procedure designed to provide 1-D predictive turbocharger models for ICE calculation. / Reyes Belmonte, MÁ. (2013). Contribution to the Experimental Characterization and 1-D Modelling of Turbochargers for IC Engines [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/34777 / TESIS

La surveillance et l'amélioration des modes de fonctionnement des systèmes propulsifs des lanceurs représentent des défis majeurs de l'industrie aérospatiale. En effet, une défaillance ou un dysfonctionnement du système propulsif peut avoir un impact significatif pour les clients institutionnels ou privés et entraîner des catastrophes environnementales ou humaines. Des systèmes de gestion de la santé (HMS) pour les moteurs fusée à ergols liquides (LPREs), ont été mis au point pour tenir compte des défis actuels en abordant les questions de sureté et de fiabilité. Leur objectif initial est de détecter les pannes ou dysfonctionnements, de les localiser et de prendre une décision à l’aide de Redlines et de systèmes experts. Cependant, ces méthodes peuvent induire de fausses alarmes ou des non-détections de pannes pouvant être critiques pour la sécurité et la fiabilité des opérations. Ainsi, les travaux actuels visent à éliminer certaines pannes critiques, mais aussi diminuer les arrêts intempestifs. Les données disponibles étant limitées, des méthodes à base de modèles sont essentiellement utilisées. La première tâche consiste à détecter les défaillances de composants et/ou d'instruments à l'aide de méthodes de détection et de localisation de fautes (FDI). Si la faute est considérée comme mineure, des actions de « non-arrêt » sont définies pour maintenir les performances de l'ensemble du système à un niveau proche de celles souhaitées et préserver les conditions de stabilité. Il est donc nécessaire d’effectuer une reconfiguration robuste (incertitudes, perturbations inconnues) du moteur. Les saturations en entrée doivent également être prises en compte dans la conception de la loi de commande, les signaux de commande étant limités en raison des caractéristiques ou performances des actionneurs physiques. Les trois objectifs de cette thèse sont donc : la modélisation des différents sous-systèmes principaux d’un LPRE, le développement d’algorithmes de FDI sur la base des modèles établis et la définition d’un système de reconfiguration du moteur en temps réel pour compenser certains types de pannes. Le système de FDI et Reconfiguration (FDIR) développé sur la base de ces trois objectifs a ensuite été validé à l’aide de simulations avec CARINS (CNES) et du banc d’essai MASCOTTE (CNES/ONERA). / Monitoring and improving the operating modes of launcher propulsion systems are major challenges in the aerospace industry. A failure or malfunction of the propulsion system can have a significant impact for institutional or private customers and results in environmental or human catastrophes. Health Management Systems (HMS) for liquid propellant rocket engines (LPREs), have been developed to take into account the current challenges by addressing safety and reliability issues. Their objective was initially to detect failures or malfunctions, isolate them and take a decision using Redlines and Expert Systems. However, those methods can induce false alarms or undetected failures that can be critical for the operation safety and reliability. Hence, current works aim at eliminating some catastrophic failures but also to mitigate benign shutdowns to non-shutdown actions. Since databases are not always sufficient to use efficiently data-based analysis methods, model-based methods are essentially used. The first task is to detect component and / or instrument failures with Fault Detection and Isolation (FDI) approaches. If the failure is minor, non-shutdown actions must be defined to maintain the overall system current performances close to the desirable ones and preserve stability conditions. For this reason, it is required to perform a robust (uncertainties, unknown disturbances) reconfiguration of the engine. Input saturation should also be considered in the control law design since unlimited control signals are not available due to physical actuators characteristics or performances. The three objectives of this thesis are therefore: the modeling of the different main subsystems of a LPRE, the development of FDI algorithms from the previously developed models and the definition of a real-time engine reconfiguration system to compensate for certain types of failures. The developed FDI and Reconfiguration (FDIR) scheme based on those three objectives has then been validated with the help of simulations with CARINS (CNES) and the MASCOTTE test bench (CNES/ONERA).

The ever increasing demands on today's engine performance and emissions control is forcing the automotive industry to make use of innovative solutions. One of these is to apply the technology of VNT turbos on commercial petrol vehicles. When using a VNT turbo, the aspect ratio of the turbine can be altered while driving to suit the current operating window. In order to actually gain performance while using a VNT, the vanes have to be properly controlled using a suitable control strategy. In this project, direct collocation have been utilized through the usage of YOP which is an adaptation of CasADi in MATLAB to solve non-linear optimization problems. Comprehensive models of the turbocharger and the cylinders have been built and validated to properly represent a VEP4 LP engine from AUROBAY. The models are implemented in YOP to create and simulate different OCPs using the turbo speed as state and position of the vanes as control signal. With this model in YOP together with the air mass flow per second as reference, a good reference following together with decent values for relevant parameters can be accomplished. Other objective functions such as minimum time and maximal volumetric efficiency are also investigated in the project which yield likewise results. From the results it can be concluded that this type of model and control strategy can be used with success when studying optimal control of a VNT turbo.

In this thesis, exergy based SI engine model optimisation (EBSIEMO) is studied and evaluated. A four-stroke bi-fuel spark ignition (SI) engine is modelled for optimisation of engine performance based upon exergy analysis. An artificial neural network (ANN) is used as an emulator to speed up the optimisation processes. Constrained particle swarm optimisation (CPSO) is employed to identify parameters such as equivalence ratio and ignition time for optimising of the engine performance, based upon maximising ¿total availability¿. In the optimisation process, the engine exhaust gases standard emission were applied including brake specific CO (BSCO) and brake specific NOx (BSNOx) as the constraints. The engine model is developed in a two-zone model, while considering the chemical synthesis of fuel, including 10 chemical species. A computer code is developed in MATLAB software to solve the equations for the prediction of temperature and pressure of the mixture in each stage (compression stroke, combustion process and expansion stroke). In addition, Intake and exhaust processes are calculated using an approximation method. This model has the ability to simulate turbulent combustion and compared to computational fluid dynamic (CFD) models it is computationally faster and efficient. The selective outputs are cylinder temperature and pressure, heat transfer, brake work, brake thermal and volumetric efficiency, brake torque, brake power (BP), brake specific fuel consumption (BSFC), brake mean effective pressure (BMEP), concentration of CO2, brake specific CO (BSCO) and brake specific NOx (BSNOx). In this model, the effect of engine speed, equivalence ratio and ignition time on performance parameters using gasoline and CNG fuels are analysed. In addition, the model is validated by experimental data using the results obtained from bi-fuel engine tests. Therefore, this engine model was capable to predict, analyse and useful for optimisation of the engine performance parameters. The exergy based four-stroke bi-fuel (CNG and gasoline) spark ignition (SI) engine model (EBSIEM) here is used for analysis of bi-fuel SI engines. Since, the first law of thermodynamic (the FLT), alone is not able to afford an appropriate comprehension into engine operations. Therefore, this thesis concentrates on the SI engine operation investigation using the developed engine model by the second law of thermodynamic (the SLT) or exergy analysis outlook (exergy based SI engine model (EBSIEM)) In this thesis, an efficient approach is presented for the prediction of total availability, brake specific CO (BSCO), brake specific NOx (BSNOx) and brake torque for bi-fuel engine (CNG and gasoline) using an artificial neural network (ANN) model based on exergy based SI engine (EBSIEM) (ANN-EBSIEM) as an emulator to speed up the optimisation processes. In the other words, the use of a well trained an ANN is ordinarily much faster than mathematical models or conventional simulation programs for prediction. The constrained particle swarm optimisation (CPSO)-EBSIEM (EBSIEMO) was capable of optimising the model parameters for the engine performance. The optimisation results based upon availability analysis (the SLT) due to analysing availability terms, specifically availability destruction (that measured engine irreversibilties) are more regarded with higher priority compared to the FLT analysis. In this thesis, exergy based SI engine model optimisation (EBSIEMO) is studied and evaluated. A four-stroke bi-fuel spark ignition (SI) engine is modelled for optimisation of engine performance based upon exergy analysis. An artificial neural network (ANN) is used as an emulator to speed up the optimisation processes. Constrained particle swarm optimisation (CPSO) is employed to identify parameters such as equivalence ratio and ignition time for optimising of the engine performance, based upon maximising ¿total availability¿. In the optimisation process, the engine exhaust gases standard emission were applied including brake specific CO (BSCO) and brake specific NOx (BSNOx) as the constraints. The engine model is developed in a two-zone model, while considering the chemical synthesis of fuel, including 10 chemical species. A computer code is developed in MATLAB software to solve the equations for the prediction of temperature and pressure of the mixture in each stage (compression stroke, combustion process and expansion stroke). In addition, Intake and exhaust processes are calculated using an approximation method. This model has the ability to simulate turbulent combustion and compared to computational fluid dynamic (CFD) models it is computationally faster and efficient. The selective outputs are cylinder temperature and pressure, heat transfer, brake work, brake thermal and volumetric efficiency, brake torque, brake power (BP), brake specific fuel consumption (BSFC), brake mean effective pressure (BMEP), concentration of CO2, brake specific CO (BSCO) and brake specific NOx (BSNOx). In this model, the effect of engine speed, equivalence ratio and ignition time on performance parameters using gasoline and CNG fuels are analysed. In addition, the model is validated by experimental data using the results obtained from bi-fuel engine tests. Therefore, this engine model was capable to predict, analyse and useful for optimisation of the engine performance parameters. The exergy based four-stroke bi-fuel (CNG and gasoline) spark ignition (SI) engine model (EBSIEM) here is used for analysis of bi-fuel SI engines. Since, the first law of thermodynamic (the FLT), alone is not able to afford an appropriate comprehension into engine operations. Therefore, this thesis concentrates on the SI engine operation investigation using the developed engine model by the second law of thermodynamic (the SLT) or exergy analysis outlook (exergy based SI engine model (EBSIEM)) In this thesis, an efficient approach is presented for the prediction of total availability, brake specific CO (BSCO), brake specific NOx (BSNOx) and brake torque for bi-fuel engine (CNG and gasoline) using an artificial neural network (ANN) model based on exergy based SI engine (EBSIEM) (ANN-EBSIEM) as an emulator to speed up the optimisation processes. In the other words, the use of a well trained an ANN is ordinarily much faster than mathematical models or conventional simulation programs for prediction. The constrained particle swarm optimisation (CPSO)-EBSIEM (EBSIEMO) was capable of optimising the model parameters for the engine performance. The optimisation results based upon availability analysis (the SLT) due to analysing availability terms, specifically availability destruction (that measured engine irreversibilties) are more regarded with higher priority compared to the FLT analysis.

[ES] Se espera que la nueva generación de motores de encendido provocado represente la mayor parte del mercado en el contexto de la propulsión de vehículos con o sin hibridación. Sin embargo, la tecnología actual todavía tiene desafíos críticos por delante para cumplir con los nuevos estándares de emisiones de CO2 y contaminantes. Consecuentemente están surgiendo nuevas tecnologías para mejorar la eficiencia de los motores y que estos cumplan con las nuevas normativas anti-contaminación. Entre otras, una de las tendencias más seguidas en la actualidad es la reducción de tamaño de los motores, concepto conocido como "downsizing", bajo la técnica de la turbosobrealimentación. Las nuevas tecnologías de turbocompresores, como las turbinas de geometría variable (TGV), se empiezan a considerar para su aplicación en las exigentes condiciones de funcionamiento de los nuevos motores de encendido provocado. En este trabajo, a partir de datos experimentales obtenidos en la sala de ensayos del motor, se propone una metodología de calibración del modelo completo de motor 1-D: se realiza un análisis teórico dirigido a asegurar el control total sobre cualquier aspecto de la simulación. En otras palabras, el modelo de motor 1-D se ajustó completamente con respecto a los datos experimentales del motor. Además, se demuestra la necesidad del postprocesamiento y validación de datos experimentales relacionados con mapas de turbocompresores, ya que se requiere desacoplar fenómenos como la transferencia de calor y las pérdidas por fricción de los denominados mapas experimentales de turbocompresores. De acuerdo con esto, se presenta una metodología para la obtención de mapas de turbocompresores, basada en una campaña experimental dividida en varias tipologias de ensayos y seguida de la etapa de modelado. La etapa de modelado se lleva a cabo utilizando modelos de turbocompresores integrales ya desarrollados o disponibles en la literatura. Adicionalmente se aborda la mejora en la precisión de las simulaciones cuando se comparan mapas de turbocompresores postprocesados con mapas puramente experimentales. Aprovechando el modelo de motor 1-D altamente validado y físicamente representativo así como los mapas validados del turbocompresor, se discute cómo las incertidumbres experimentales o las variables "fuera de control" pueden afectar los resultados experimentales. Se propone una metodología para superar este punto desde la perspectiva del modelado. Lo anterior permite realizar comparativas que en las se analiza exclusivamente el impacto de diferentes tecnologías de turbina o unidades de turbinas. Además, tomando como base el modelo ya desarrollado, es posible explorar diferentes cálculos de optimización, estrategias de control y proporcionar comparaciones de tecnología de turbinas en plenas cargas y cargas parciales de motor en un amplio rango de revoluciones. También se aborda el impacto de la altitud y se evalúan los transitorios de carga para dos tecnologías de turbinas analizadas: VGT y WG. Como conclusión, se demuestra que la tecnología VGT muestra menos limitaciones en condiciones de trabajo extremas, como en la curva de plena carga, donde la tecnología WG representa una limitación en términos de máxima potencia. Las diferencias a plena carga se vuelven aún más evidentes en condiciones de trabajo en altitud. Cuando se trata de cargas parciales, las diferencias en el consumo de combustible son menores, pero potencialmente beneficiosas para los VGT. / [CA] S'espera que la nova generació de motors d'encesa per espurna representi la major part del mercat en el context de la propulsió de vehicles amb o sense hibridació. No obstant això, la tecnologia actual encara té reptes crítics per davant per complir amb els nous estàndards d'emissions de CO2 i contaminants. Conseqüentment estan sorgint noves tecnologies per millorar l'eficiència dels motors i que aquests compleixin amb les noves normatives anti-contaminació. Entre d'altres, una de les tendències més seguides en l'actualitat és la reducció de grandària dels motors, concepte conegut com "downsizing", sota la tècnica de la turbosobrealimentación. Les noves tecnologies de turbocompressors, com les VGT, es comencen a considerar per la seva aplicació en les exigents condicions de funcionament dels nous motors d'encesa per espurna. En aquest treball, a partir de dades experimentals obtingudes a la sala d'assajos de l'motor, es proposa una metodologia de calibratge del model complet de motor 1-D: es realitza una anàlisi teòrica dirigit a assegurar el control total sobre qualsevol aspecte de la simulació. En altres paraules, el model de motor 1-D es va ajustar completament respecte a les dades experimentals del motor. A més, es demostra la necessitat del posprocesamiento i validació de dades experimentals relacionats amb mapes de turbocompressors, ja que es requereix desacoblar fenòmens com la transferència de calor i les pèrdues per fricció dels denominats mapes experimentals de turbocompressors. D'acord amb això, es presenta una metodologia per a l'obtenció de mapes de turbocompressors, basada en una campanya experimental dividida en diverses tipologies d'assajos i seguida de l'etapa de modelatge. L'etapa de modelatge es porta a terme utilitzant models de turbocompressors integrals ja desenvolupats disponibles a la literatura. A més a s'aborda la millora en la precisió de les simulacions quan es comparen mapes de turbocompressors postprocessats amb mapes purament experimentals. Aprofitant el model de motor 1-D validat i físicament representatiu així com els mapes validats del turbocompressor, es discuteix com les incerteses experimentals o les variables "fora de control" poden afectar els resultats experimentals. Es proposa una metodologia per superar aquest punt des de la perspectiva de la modelització. L'anterior permet realitzar exclusivament la comparació de tecnologies / unitats de turbines. A més, prenent com a base el model ja desenvolupat, és possible explorar diferents càlculs d'optimització, estratègies de control i proporcionar comparacions de tecnologia de turbines a càrregues completes i parcials del motor en un ampli rang de revolucions del motor. També s'aborda l'impacte de l'altitud i s'avaluen els transitoris de càrrega per a dues tecnologies de turbines analitzades: VGT i WG. com a conclusió, es demostra que la tecnologia VGT mostra menys limitacions en condicions de treball extremes, com en la corba de plena càrrega, on la tecnologia WG representa una limitació en termes de màxima potència. Les diferències a plena càrrega es tornen encara més evidents en condicions de treball en altitud. Quan es tracta de càrregues parcials, les diferències en el consum de combustible són menors, però potencialment beneficioses per als VGT. / [EN] The new generation of spark ignition (SI) engines is expected to represent most of the future market share in the context of power-train with or without hybridization. Nevertheless, the current technology has still critical challenges in front to meet incoming CO2 and pollutant emissions standards. Consequently, new technologies are emerging to improve engine efficiency and meet new pollutant regulations. Among others, one of the most followed trends is engine size reduction, known as downsizing, based on the turbocharging technique. New turbocharger technologies, such as variable geometry turbines (VGT), are evaluated for their application under the demanding operating conditions of SI engines. In this work, from experimental data obtained in an engine test cell, a 1-D complete engine model calibration methodology was conducted: a theoretical analysis aimed at ensuring full control on any aspect of the simulation. In other words, the 1-D engine model was fully fitted with respect to the experimental engine data. Furthermore, it is evidenced the requirement of post-processing and validating the experimental data dealing with turbocharger maps, since phenomena such as heat transfer and friction losses are required to be decoupled from the so-called experimental turbocharger maps. Accordingly, a methodology for turbocharger maps obtention is presented, based on an experimental campaign divided into several test typologies and followed by the modelling stage. The modelling stage is carried out making usage of already developed integral turbocharger models available in the literature. Additionally, the improvement in the accuracy of the simulations when post-processed turbocharger maps are compared against purely experimental maps is addressed. Taking advantage of the highly validated and physically representative 1-D gas-dynamics engine model and turbocharger validated maps, it is discussed how experimental uncertainties or "out-of-control" variables may impact the experimental results. A methodology is proposed to overcome this point from the modelling perspective. The previous allows performing exclusively turbine technologies/units comparison. In addition, taking as a basis the already developed model, it is possible to explore different optimization calculations, control strategies and provide turbine technology comparisons at engine full and partial loads in a wide range of engine speed. Also, the altitude impact is addressed and load transients are evaluated for two analysed turbine technologies: VGT and WG. In all, it was found that VGT technology shows fewer limitations in extreme working conditions, such as full load curve, where the WG technology represents a limitation in terms of the maximum power output. Full load differences become even more evident in altitude working conditions. When it comes to partial loads, differences in fuel consumption are minor but potentially beneficial for VGTs. / Gómez Vilanova, A. (2022). Modelling and analysis methodology of SI IC engines turbocharged by VGT [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/181929 / TESIS

Developing and improving upon a good empirical model for an engine can be time-consuming and costly. The goal of this thesis has been to evaluate data-driven modelling, specifically neural networks, to see how well it can handle training for some static models like the mass flow of air into the cylinder, mean effective pressure and pump mean effective pressure but also for transient modelling, specifically the exhaust gas temperature. These models are evaluated against the classical empirical models to see if neural networks are a viable modelling option. This is done with five different types of neural networks which are trained. These are the feed-forward neural network, Nonlinear autoregressive exogenous model network, layer recurrent network, long short term memory network and gated recurrent network.The inputs were determined by looking at more simple physical models but also looking at the covariance to determine the usefulness of the input. If the calculation time is small for the specific network, the neural network structure is tested and optimized by training many networks and finding the median/mean result for that specific test.The result has shown that the static models are handled very well by the most simple feed-forward network. For the exhaust temperature, both NARX and Layer recurrent network could predict and handle it well giving results very close to the empirical models and could be a viable option for transient modelling, on the other hand, Long short term memory, gated recurrent network and the feed-forward network had trouble predicting the exhaust gas temperature and returned bad results while training.