Methodology for underhood air prediction for low fidelity computational models

Thermal management is critical for modern vehicles, particularly for Zero Emission Vehicles (ZEVs), where maintaining optimal temperature ranges directly influences thermal system efficiency and vehicle range. Accurate prediction of underhood airflow behavior is essential for effective thermal management and also to estimate overall energy consumption by cooling system, with air-side dynamics playing a pivotal role in heat transfer over the heat exchangers of cooling package. Simulation tools like GT-Suite are indispensable for this purpose, enabling engineers to evaluate complex thermal interactions without the cost and time constraints of extensive physical testing. While 3D Computational Fluid Dynamics (CFD) models offer detailed insights into flow characteristics, they are computationally expensive and time consuming. In contrast, 1D models provide faster simulation times, making them ideal for system-level analysis and iterative design processes. However, 1D models inherently lack the ability to capture detailed flow phenomena, which can compromise the accuracy of thermal predictions. To mitigate this, calibration using 3D CFD data or experimental measurements becomes critical, ensuring that air-side behavior is represented as accurately as possible. One of the key challenges in this calibration process arises at low fan speeds, where matching flow rates becomes difficult due to the unavailability of windmilling data. Along with calibration of normal operating conditions, this paper also presents a methodology for tuning fan maps under such constraints, focusing on strategies to enhance model fidelity and novel methodology to calculate fan mechanical power. We explore simulation-based techniques, leveraging steady-state operating conditions to refine fan characteristics. The study further discusses sensitivity analysis, validation strategies, and potential inaccuracies introduced by missing windmilling effects and method to accurately fill in the missing fan map data. The proposed methodology ensures improved predictive accuracy of underhood airflow behavior, enhancing thermal system design for automotive applications. This method improves the reliability of underhood airflow predictions, ultimately contributing to more accurate thermal management system predictions out of digital tools.