This content is not included in your SAE MOBILUS subscription, or you are not logged in.
Development and Experimental Validation of Empirical and Semi-Physical Exhaust Gas Temperature Models
ISSN: 0148-7191, e-ISSN: 2688-3627
To be published on September 27, 2020 by SAE International in United States
Exhaust gas temperature is one of the main parameters under the focus of engine manufacturers, due to its effects on both Turbo-Charger durability and catalyst efficiency. Typically, the measurement of such variable at the test bench is carried out by using thermocouples that are not available for on-board application. For this reason, an accurate and reliable model for Real-Time calculation of this variable is particularly important, to develop more precise and efficient strategies to prevent damages to the exhaust system and optimize catalyst efficiency. In this work, two different models for the calculation of the exhaust gas temperature are described and validated, comparing the model outputs with experimental measurements. At first a control-oriented empirical model is presented, describing how the polynomial approach used in a previous work of the authors has been applied to reproduce the steady-state thermocouple measurement. Moreover, a Real-Time compatible thermocouple dynamic model is introduced, and the calculated values are then compared with the thermocouple signal under dynamic conditions. A simulation-oriented semi-physical Wiebe-based model that calculates the combustion Rate of Heat Release is then discussed. The main contribution is the identification process of its main parameters and the definition of the transfer function that converts the temperature of the gas in the combustion chamber at the Exhaust-Valve-Opening into the value measured under steady-state conditions in the exhaust manifold. The resulting temperature is then used as input for the thermocouple dynamics model, and the results are compared with the measured signals in transient conditions. In the last section of the paper both models are finally compared in terms of accuracy and computational effort, highlighting the more suitable application for each presented approach.