Search Results

Energy management of battery electric vehicle (BEV) is a very important and complex multi-system optimisation problem. The thermal energy management of a BEV plays a crucial role in consistent efficiency and performance of vehicle in all weather conditions. But in order to manage the thermal management, it requires a significant number of temperature sensors throughout the car including high voltage batteries, thus increasing the cost, complexity and weight of the car. Virtual sensors can replace physical sensors with a data-driven, physical relation-driven or machine learning-based prediction approach. This paper presents a framework for the development of a neural network virtual sensor using a thermal system hardware-in-the-loop test rig as the target system. The various neural network topologies, including RNN, LSTM, GRU, and CNN, are evaluated to determine the most effective approach.

Ammonia has received attention as an alternative hydrogen carrier and a potential fuel for thermal propulsion systems with a lower carbon footprint. One strategy for high power density in ammonia applications will be direct injection of liquid ammonia. Understanding the evaporation and mixing processes associated with this is important for model development. Additionally, as a prior step for developing new injectors, it is of interest to understand how a conventional gasoline direct injection (GDI) injector would behave when used for liquid ammonia without any modifications. Pure anhydrous ammonia, in its liquid form, was injected from a single hole GDI injector at a fuel pressure of 150 bar into an optically accessible constant volume chamber filled with nitrogen gas for ammonia spray measurements. The chamber conditions spanned a wide range of pressures from 3 − 15 bar at an increment of 1 bar or 2 bar between the test points.

Several governments are increasing the blending mandate of renewable fuels to reduce the life-cycle greenhouse gas emissions of the road transport sector. Currently, ethanol is a prominent renewable fuel and is used in low-level blends, such as E10 (10 %v/v ethanol, 90 %v/v gasoline) in many parts of the world. However, the exact concentration of ethanol amongst other renewable fuel components in commercially available fuels can vary and is not known. To understand the impact of the renewable fuel content on the emissions from Euro 6d-TEMP emissions specification vehicles, this paper examines the real-driving emissions (RDE) from four 2020 to 2022 model-year vehicles run on E0 and E10 fuels. CO, CO2, NO, and NO2 were measured through a Portable Emissions Measuring System (PEMS).

Forthcoming worldwide emissions regulations will start regulating ammonia emissions from light duty vehicles. At present, most light duty vehicles are powered by gasoline spark ignition engines. Sources of ammonia emission from such engines can be in-cylinder reactions (i.e. combustion) or downstream reactions across aftertreatment devices, particularly three-way catalysts. The latter has been known to be a major source of ammonia emissions from gasoline vehicles and has been extensively investigated. The former (combustion), less so, and thus is the subject of this work. A two-zone thermodynamic spark ignition engine model with a comprehensive chemical kinetics framework (C3MechV3.3 mechanism), after being validated against experimental ammonia emissions data, is used to study ammonia formation during combustion.

As lithium-ion electric vehicle (EV) batteries are sensitive to the conditions they are exposed to during charging and discharging, operational control has been an important research area. While an understanding of the effects current load and operation temperature has on the ageing stability of a battery has been established, associated control strategies are yet to be fully optimized. Most battery charging studies utilize controlled ambient temperatures and basic defined cycles, which may only apply to a small subset of real-world EV consumers. This leads to control strategies that do not consider electrical grid price fluctuation, user driving habits or local weather conditions. This paper looks to propose improved smart charging strategies of EVs to reduce consumer costs while also increasing the battery longevity. To accomplish the primary objective, A model has been generated that simulates the standard charge cycle of a battery.

GTDI engines are becoming more efficient, whether individually or part of a HEV (Hybrid Electric Vehicle) powertrain. For the latter, this efficiency manifests itself as increase in zero emissions vehicle mileage. An ideal device for energy recovery is a turbogenerator (TG), and, when placed downstream the conventional turbine, it has minimal impact on catalyst light-off and can be used as a bolt-on aftermarket device. A Ricardo WAVE model of a representative GTDI engine was adapted to include a TG (Turbogenerator) and TBV (Turbine Bypass Valve) with the TG in a mechanical turbocompounding configuration, calibrated using steady state mapping data. This was integrated into a co-simulation environment with a SISO (Single-Input, Single-Output) dynamic controller developed in SIMULINK for the actuator control (with BMEP, manifold air pressure and TG pressure ratio as the controlled variables).

In modern powertrains systems, sensors are critical elements for advanced control. The identification of sensing requirements for such highly nonlinear systems is technically challenging. To support the sensor selection process, this paper proposes a methodology to quantify the information gained from sensors used to control nonlinear dynamic systems using a dynamic probabilistic framework. This builds on previous work to design a Bayesian observer to deal with nonlinear systems. This was applied to a bimodal model of the SCR aftertreatment system. Despite correctly observing the bimodal distribution of the internal Ammonia-NOx Ratio (ANR) state, it could not distinguish which state is the true state. This causes issues for a control engineer who is less interested in how precise a measurement is and more interested in the location within control parameter space. Information regarding the dynamics of the systems is required to resolve the bimodality.

To improve the state of the art in automotive aerodynamic prediction using CFD, it is important to compare different CFD methods, software and modelling for standardized test cases. This paper reports on the 2nd Automotive CFD Prediction Workshop for the Windsor body squareback test case. The Windsor model has high quality experimental data available and a simple geometry that allows it to be simulated with limited computational resources. The model is 1 metre long and operates at a Reynolds number of 2.7 million. The original Windsor model did not include wheels, but a second variant was added here with non-rotating wheels. Experimental data is available for integrated forces, surface pressure and wake PIV surveys. Eight standard meshes were provided, covering the two geometry variants, two near wall mesh spacings (relating to wall resolved and wall modelled) and two mesh densities in the wake (relating to RANS and eddy resolving).

Particulate emissions from gasoline direct injection (GDI) engines continue to be a topic of substantial research interest. Forthcoming regulation both in the USA and the EU will further reduce their emission and drive innovation. Substantial research effort is spent undertaking experiments to understand, characterize, and research particle number (PN) emissions from engines and vehicles. Recent advances in computing power, data storage, and understanding of artificial intelligence algorithms now mean that these are becoming an important tool in engine research. In this work a random forest (RF) algorithm is used for the prediction of PN emissions from a highly boosted (up to 32 bar BMEP) GDI engine. Particle size, concentration, and the accumulation mode geometric standard deviation (GSD) are all predicted by the model. The results are analysed and an in depth study on parameter importance is carried out.

Low temperature combustion (LTC) of diesel fuel offers a path to low engine emissions of nitrogen oxides (NOx) and particulate matter (PM), especially at low loads. Borescopic optical imaging offers insight into key aspects of the combustion process without significantly disrupting the engine geometry. To assess LTC combustion, two-colour pyrometry can be used to quantify local temperatures and soot concentrations (KL factor). High sensitivity photo-multiplier tubes (PMTs) can resolve natural luminosity down to low temperatures with adequate signal-to-noise ratios. In this work the authors present the calibration and implementation of a borescope-based system for evaluating low luminosity LTC using spatially resolved visible flame imaging and high-sensitivity PMT data to quantify the luminous-area average temperature and soot concentration for temperatures from 1350-2600 K.

A model-based urea-dosing controller has been developed for the selective catalytic reduction (SCR) units on a diesel engine exhaust aftertreatment system (EATS). The SCR units consist of an integrated SCR-coated filter and then followed by a flow-through SCR catalyst. The controller was developed based on an analysis of the data generated from a Millbrook London Transport Bus (MLTB) test cycle fed into a validated model of the SCR-filter and SCR units. The critical system parameters that showed strong correlation with outlet nitrogen oxides (NOx) and ammonia (NH3) emissions were first identified, and then the sensitivity of those parameters was analyzed. The most sensitive system parameters were configured as the controller gain parameters. A proportional controller based on the key parameters with optimized gains settings for the MLTB cycle delivered over a 10% reduction in cumulative NOx emission over the cycle compared to a fixed NH3/NOx ratio (ANR) controller.

Full hybrid electric vehicles are usually defined by their capability to drive in a fully electric mode, offering the advantage that they do not produce any emissions at the point of use. This is particularly important in built up areas, where localized emissions in the form of NOx and particulate matter may worsen health issues such as respiratory disease. However, high degrees of electrification also mean that waste heat from the internal combustion engine is often not available for heating the cabin and for maintaining the temperature of the powertrain and emissions control system. If not managed properly, this can result in increased fuel consumption, exhaust emissions, and reduced electric-only range at moderately high or low ambient temperatures negating many of the benefits of the electrification. This paper describes the development of a holistic, modular vehicle model designed for development of an integrated thermal energy management strategy.

Significant aerodynamic drag reduction is obtained on a bluff body by tapering the rear body. In the 1930’s it was found that a practical low drag car body could be achieved by cutting off the tail of a streamlined shape. The rear end of a car with a truncated tail is commonly referred to as a Kamm back. It has often been interpreted as implying that the drag of this type of body is almost the same as that for a fully streamlined shape. From a review of the limited research into truncated streamlined tails it is shown in this paper that, while true for some near axisymmetric bodies, it is not the case for many more car-like shapes. For these shapes the drag reduction from an elongated tail varies almost linearly with the reduction in cross section area. A CFD simulation to determine the drag reduction from a truncated streamlined tail of variable length on the simple Windsor Body is shown by way of confirmation.

Predicting Exterior Water Management is important for developing vehicles that meet customer expectations in adverse weather. Fluid film methods, with Lagrangian tracking, can provide spray and surface water simulations for complex vehicle geometries in on-road conditions. To cope with this complexity and provide practical engineering simulations, such methods rely on empirical sub-models to predict phenomena such as the film stripping from the surface. Experimental data to develop and validate such models is difficult to obtain therefore here a high-fidelity Coupled Level-set Volume of Fluid (CLSVOF) simulation is carried out. CLSVOF resolves the interface of the liquid in three dimensions; allowing direct simulation of film behaviour and interaction with the surrounding air. This is used to simulate a simplified wing-mirror, with air flow, on which water is introduced.

Proton exchange membrane fuel cell (PEMFC) provides a promising future low carbon automotive powertrain solution. The catalyst layer (CL) is its core component which directly influences the output performance. PEMFC performance can be greatly improved by the effective optimization of CL composition. This work demonstrates a deep optimization of CL composition for improving the PEMFC performance, including the platinum (Pt) loading, Pt percentage of carbon-supported Pt and ionomer to carbon ratio of the anode and the cathode,. The simulation results by a PEMFC three-dimensional (3D) computation fluid dynamics (CFD) model coupled with the CL agglomerate model is used to train the artificial neural network (ANN) which can efficiently predict the current density under different CL composition. Squared correlation coefficient (R-square) and mean percentage error in the training set and validation set are 0.9867, 0.2635% and 0.9543, 1.1275%, respectively.

The European Particle Measurement Program (PMP) defines the current standard for measurement of Particle Number (PN) emissions from vehicles in Europe. This specifies a 50% count efficiency (D50) at 23 nm and a 90% count efficiency (D90) at 41 nm. Particulate emissions from Gasoline Direct Injection (GDI) engines have been widely studied, but usually only in the context of PMP or similar sampling procedures. There is increasing interest in the smallest particles - i.e. smaller than 23 nm - which can be emitted from vehicles. The literature suggest that by moving D50 to 10 nm, PN emissions from GDI engines might increase by between 35 and 50% but there remains a lot of uncertainty.

The partial melting of ingested ice crystals can lead to ice accretion in aircraft compressors, but accurately measuring the relatively small fraction of liquid water content in such flows is challenging. Probe-based methods for detecting liquid water content are not suitable for deployment within turbofan engines, and thus alternatives are sought. Recent research has described approaches based on passive microwave sensing. We present here an approach based on active microwave transmission and reflection, employing a vector network analyzer. Utilization of both transmission and reflection provides additional data over and above emission or transmission only, and permits a more controllable environment than passive sensing approaches. The paper specifically addresses the question of whether such an approach is viable within the context of representative icing wind tunnel and engine flow conditions.

Ingestion of high altitude atmospheric ice particles can be hazardous to gas turbine engines in flight. Ice accretion may occur in the core compression system, leading to blockage of the core gas path, blade damage and/or flameout. Numerous engine powerloss events since 1990 have been attributed to this mechanism. An expansion in engine certification requirements to incorporate ice crystal conditions has spurred efforts to develop analytical models for phenomenon, as a method of demonstrating safe operation. A necessary component of a complete analytical icing model is a thermodynamic accretion model. Continuity and energy balances are performed using the local flow conditions and the mass fluxes of ice and water that are incident on a surface to predict the accretion growth rate.

A significant number of historical engine powerloss events have recently been attributed to ingestion of high altitude ice crystals, prompting regulators to expand engine certification envelopes to incorporate ‘ice crystal icing’ conditions. There has been a resulting effort by OEMs and academia to develop analytical and semi-empirical models for the phenomenon, partly through use of rig testing. The current study presents results and analysis of experiments conducted in the National Research Council’s Research Altitude Test Facility (RATFac). The experiments used a simplified compressor stator vane test article, designed to produce data to build semi-empirical models and validate an existing ice crystal icing code. Accretion growth rates, extracted from backlit shadowgraphy, are presented as a function of test condition, and the algorithm of a new image processing technique using Canny filtering is discussed.

Numerous turbofan power-loss events have occurred in high altitude locations in the presence of ice crystals. It is theorized that ice crystals enter the engine core, partially melt in the compressor and then accrete onto stator blade surfaces. This may lead to engine rollback, or shed induced blade damage, surge and/or flameout. The first generation of ice crystal icing predictive models use a single flow field where there is no accretion to calculate particle trajectories and accretion growth rates. Recent work completed at the University of Oxford has created an algorithm to automatically detect the edge of accretion from experimental video data. Using these accretion profiles, numerical simulations were carried out at discrete points in time using a manual meshing process.