1D Simulation Accuracy Enhancement for Predicting Powertrain Cooling System Performance 2019-26-0298

In today’s competitive scenario, the automotive product life cycle has drastically reduced and all Auto OEM’s are coming up with their updated products with lesser development time. These frequent product upgrades are possible due to use of various digital tools during product design and development. Design and optimization of engine coolpack (powertrain cooling unit) to attain engine cooling performance is one of the important parameter during vehicle development or upgrade. Hence, to keep control over development cost and time of delivery, quick and accurate digital validation capability like one dimensional (1D) simulation is the need of the hour.

To predict the powertrain cooling (PTC) performance at vehicle concept stage, when physical prototypes are not available, airflow data from similar developed platforms is considered as an input for 1D simulation. As the vehicle program matures and the 3D CFD airflow simulation results become available as an input, 1D simulation accuracy of 87~90% can be achieved. However, even at this stage, 3D CFD airflow simulation results are provided under the assumption of isothermal condition. Due to this assumption, the effect of heat rejection from upstream components in the coolpack (consisting of condenser, charge air cooler, radiator and fan) on airflow is not taken into account. In addition, the proximity of the coolpack to the engine may result in hot air recirculation. Additionally, the heat rejection and coolant flow rate used for simulation measured at engine test bed differ from that experienced in actual vehicle.

This paper addresses the effect of heat rejection from upstream components on the density of air as it moves across the coolpack, the effect of hot air recirculation and the correction factors required in inputs obtained from engine test bed. The methodology and model developed in this work have been validated against real world physical test data. Using this proposed methodology simulation prediction with an accuracy of 95~97% is obtained.