1D Simulation Accuracy Enhancement for Predicting Powertrain Cooling System Performance 2019-26-0298
In today’s vehicle product development scenario, cost involved in physical testing during development stage is very high. To predict the Powertrain Cooling (PTC) performance early at vehicle conceptualization stage when physical prototype are not available, airflow from previous similar platform is considered as input. As the program matures, by using airflow provided by 3D CFD, 1D simulation accuracy can be obtained upto 90%. At this stage, the airflow over components is provided at isothermal conditions by 3D CFD and is directly taken as input for 1D CAE. Due to this, the effect of heat rejection of upstream components on airflow is not taken into account during simulation. This paper addresses the effect of heat rejection from upstream components on the density of air as it moves across the cool-pack components so as to deliver more reliable simulation predictions for early risk assessment.
PTC performance is strongly dependent on cool-pack airflow and is a critical input data for simulations. Small deviation in airflow would lead to over/under prediction of cool pack performance. Even ±5% deviation in airflow (m/s) effects Upper Coolant Line (UCL) temperature (°C) by ~4%. The density of air changes as it flows across cool-pack components due to heat rejection by different heat exchangers. To achieve accurate PTC performance prediction, effect of change in density needs to be modelled.
The revised approach considering density takes the airflow and static pressure values before and after different heat exchangers from 3D CFD as input for 1D simulation. The static pressure values are used to calculate coefficient of pressure (Cp) and the build in resistance (BiR). Then, isothermal coolant and air circuit is modelled and is calibrated to achieve airflow predicted by 3D CFD. Thereafter, similar to conventional process the final simulation is carried out by considering heat rejection from various components and its effect on air density which in-turn corrects the airflow.
The revised approach considering density approach helps to improve prediction accuracy from ~90 to 97%. Hence, we can rely on simulation results and cut down physical test iterations on various vehicles at different operating conditions.