Search Results

In this study, the combustion system of a light-duty compression ignition engine running on a market gasoline fuel with Research Octane Number (RON) of 91 was optimized using computational fluid dynamics (CFD) and Machine Learning (ML). This work was focused on optimizing the piston bowl geometry at two compression ratios (CR) (17 and 18:1) and this exercise was carried out at full-load conditions (20 bar indicated mean effective pressure, IMEP). First, a limited manual piston design optimization was performed for CR 17:1, where a couple of pistons were designed and tested. Thereafter, a CFD design of experiments (DoE) optimization was performed where CAESES, a commercial software tool, was used to automatically perturb key bowl design parameters and CONVERGE software was utilized to perform the CFD simulations. At each compression ratio, 128 piston bowl designs were evaluated.

The current research focuses in understanding how inlet pressure, EGR, combustion phasing, engine speed and pilot main ratio are affecting the main parameters of the combustion (e.g. efficiency, NOx, soot, maximum pressure rise rate) in the novel concept of injecting high octane number fuels in partially premixed combustion. The influence of the above mentioned parameters was studied by performing detailed sweeps at 32 bar fuel MEP (c.a. 16-18 bar gross IMEP); three different kinds of gasoline were tested (RON: 99, 89 and 69). The experiments were ran in a single cylinder heavy duty engine; Scania D12. At the end of these sweeps the optimized settings were computed in order to understand how to achieve high efficiency, low emissions and acceptable maximum pressure rise rate.

This paper describes the development of a novel data model for storing and sharing data obtained from engine experiments, it then outlines a methodology for automatic model development and applies it to a state-of-the-art engine combustion model (including chemical kinetics) to reduce corresponding model parameter uncertainties with respect engine experiments. These challenges are met by adopting the latest developments in the semantic web to create a shared data model resource for the IC engine development community. The relevant data can be extracted and then used to set-up simulations for parameter estimation by passing it to the relevant application models. A methodology for incorporating experimental and model uncertainties into the model optimization procedure is presented.

A previously presented robust and fast diagnostic NOx model was modified into a predictive model. This was done by using simple yet physically-based models for fuel injection, ignition delay, premixed heat release rate and diffusion combustion heat release rate. The model can be used both for traditional high temperature combustion and for high-EGR low temperature combustion. It was possible to maintain a high accuracy and calculation speed of the NOx model itself. The root mean square of the relative model error is 16 % and the calculation speed is around one second on a PC. Combustion characteristics such as ignition delay, CA50 and the general shape of the heat release rate are well predicted by the combustion model. The model is aimed at real time NOx calculation and optimization in a vehicle on the road.

In this paper a fast NOx model is presented which can be used for engine optimization, aftertreatment control or virtual mapping. A cylinder pressure trace is required as input data. High calculation speed is obtained by using table interpolation to calculate equilibrium temperatures and species concentrations. Test data from a single-cylinder engine and from a complete six-cylinder engine have been used for calibration and validation of the model. The model produces results of good agreement with emission measurements using approximately 50 combustion product zones and a calculation time of one second per engine cycle. Different compression ratios, EGR rates, injection timing, inlet pressures etc. were used in the validation tests.

The effects of intake pressure (Pin), start of injection (SOI), injection pressure (Pinj), injection split ratio (Qsplit), internal and external exhaust gas recirculation rates were varied to optimize several key parameters at a partially pre-mixed combustion low load/low speed condition using CFD. These include indicated specific fuel consumption (ISFC), combustion phasing (CA50), maximum rate of pressure rise (MRPR), maximum cylinder pressure (Pmax), indicated specific NOx (sNOx), indicated specific hydrocarbons (sHC) and Filter Smoke Number (FSN) emissions. Low-load point (6 bar indicated mean effective pressure (IMEP), 1500 revolutions per minute (RPM)) was selected where Pin varied between 1.25 and 1.5 bar, SOI between -100 and -10 crank angle degree (CAD) after top dead center (aTDC), Pinj between 100 and 200 bar, split ratio between 0 and 0.5, EGR between 0 and 45% and internal EGR measured by rebreathing valve height was varied between 0 and 4.5 mm.