Automotive manufacturers nowadays are constantly working on improving their internal combustion engines’ performance by reducing the fuel consumption and emissions, without compromising the power generated. Manufacturers are therefore relying on virtual engine models that can be run on simulation software in order to reduce the amount of time and costs needed, in comparison with experiments done on engine test benches. One important element of the intake system of an internal combustion engine is the throttle valve, which defines the amount of air reaching the plenum before being drawn into the cylinders. This article discusses a widely used model for the estimation of air flow rate through the throttle valve in an internal combustion engine simulation. Experiments have been conducted on an isolated throttle valve test bench in order to understand the influence of different factors on the model’s discharge coefficient. These experiments showed that the discharge coefficient varies with the pressure ratio across the throttle valve and with its angle. Furthermore, for each angle, this variation can be approximated with a linear model composed of two parameters: the slope and the Y-Intercept. These parameters are calibrated for different throttle valve angles. This calibration can be done using automotive manufacturers’ standard engine test fields that are often available. This model is then introduced into an engine simulation model, and the results are compared to the experimental data of a turbocharged engine test bench for validation. They are also compared with a standard discharge coefficient model that varies only with the throttle valve angle. The results show that the new model for the discharge coefficient reduces mass flow estimation errors and allows expanding the applications of the throttle valve isentropic nozzle model.
Teaching Nonlinear Model Predictive Control with MATLAB/Simulink and an Internal Combustion Engine Test Bench
