Search Results

The development of the Diesel Particulate Filter (DPF) cell geometry and DPF size for new applications requires specific tools to predict the pressure drop as a function of filter characteristics, mass flow and filter loading. A 1-D permeability model is most useful for this type of work. This paper presents the development of a 1-D physical model of DPF permeability. This model includes the symmetric and asymmetric channel shape and is able to simulate various functional phases of the DPF through its lifetime: with or without soot and with or without ash. This kind of model needs several physical coefficients, in order to describe the flow behavior. This work explains the determination of the physical coefficients of the 1-D model. The large disparity of the literature is shown. Therefore, it is necessary to carefully determine these coefficients.

This paper presents a new 0D phenomenological approach to predict the combustion process in diesel engines operated under various running conditions. The aim of this work is to develop a physical approach in order to improve the prediction of in-cylinder pressure and heat release. The main contribution of this study is the modeling of the premixed part of the diesel combustion with a further extension of the model for multi-injection strategies. In phenomenological diesel combustion models, the premixed combustion phase is usually modeled by the propagation of a turbulent flame front. However, experimental studies have shown that this phase of diesel combustion is actually a rapid combustion of part of the fuel injected and mixed with the surrounding gas. This mixture burns quasi instantaneously when favorable thermodynamic conditions are locally reached. A chemical process then controls this combustion.

This paper describes an innovative method to estimate the wall losses during the compression and combustion strokes of a gasoline engine using the cylinder pressure measurement. The estimation during the compression and combustion strokes allows to better represent the system during the combustion. A sliding mode observer is derived from a validated 0-D physical engine model and its convergence and stability are proved. The observer is validated using two different engine models: a one zone engine model and a two zones engine model with flame wall interaction. A good agreement between the estimation results and the model reference is observed, showing the interest of using closed loop strategies to estimate the wall losses in a SI engine.