Diesel engine out exhaust temperature modelling- moving from a map based to physically based model approach

Abstract

Engine out exhaust gas is an input to the engine system components downstream, which are directly or indirectly responsible for air charge control, emission regulation, torque control or are used as feedback to the engine control system. This makes the accurate prediction of the exhaust gas an important criterion for the control system. Automotive manufacturers use map based models for the prediction of exhaust temperature at known engine operating conditions. The aim of this thesis is to design a predictive physically based exhaust temperature model for a typical diesel engine. The model predicts the temperature of the exhaust gases at the end of the exhaust port with acceptable levels of accuracy. Previously available map based models were specific to the engine and required extensive test data for creation and validation of such maps. The model developed in this thesis is physically based, which enables its usage across a range of typical diesel engines with reasonable modifications. This thesis also includes two different modelling approaches for the temperature prediction. One being an averaged out cyclic approach while the other being a more resolved computation at crank angle degrees. Data from test rig is used as input parameter for calibrating and tuning the models. An attempt to understand the transient effects at the end of the exhaust port is also a part of the thesis and precedes the discussion on modelling.