CFD optimization of ventilation ducts and panel outlets for a passenger vehicle

Andrea Alessandro Piovano, Laura Lorefice, Nicola Paola, Enrico Ribaldone, Luca Miretti

Fiat Group Automobiles


The ventilation system in a passenger vehicle is designed to keep the occupants comfortable delivering air at the correct velocity and temperature levels. In order to design such system, before physical prototypes testing, virtual simulations are required, allowing to predict accurately the internal flow through ducts and cabin interior. The use of CFD at the earliest stages of the project is very important, as late changes in prototypes phase are expensive and time-consuming. This developed methodology will reduce the late changes on a project by more than 30%.
The CFD simulation model allows to evaluate how well the A/C system performs in terms of directing the airflow on the front driver and passenger, giving a quantitative prediction of airflow distribution pattern respect to specific location on plywood/manikin man. The CFD optimization has been conducted in order to find the optimum ducts shape, that permits to reach the lowest pressure drop, and the primary and secondary vanes tilt angles, that allow to reach the velocity targets on a plywood, for different aiming setups. The objectives to optimize were identified in ducts pressure drop and velocity of the target areas around the structure of the plywood, representing the different parts of the human body: eyes, shoulders, chest and elbows. These results allow to understand if the tilt angles limits, panel outlet and ducts geometry must be changed, before the physical prototypes testing, in the earliest phase of the project. Starting from the modelling of the complete cabin of a B-SUV vehicle, the parametrization of the vanes rotation axis and angles has been set up, while for the ducts three geometrical shape features have been considered. After that, two different methods for multi-objective optimization were evaluated and the most efficient one was chosen.
The fluid dynamics simulation has been carried out using STAR-CCM+, while the optimization process has been driven by modeFRONTIER.