Simmonds, N., Tsoutsanis, P., Drikakis, D., Gaylard, A. et al., "Full Vehicle Aero-Thermal Cooling Drag Sensitivity Analysis for Various Radiator Pressure Drops," SAE Technical Paper 2016-01-1578, 2016, doi:10.4271/2016-01-1578.
Simulations are presented which fully couple both the aerodynamics and cooling flow for a model of a fully engineered production saloon car (Jaguar XJ) with a two-tier cooling pack. This allows for the investigation of the overall aerodynamic impact of the under-hood cooling flow, which is difficult to predict experimentally.The simulations use a 100 million-element mesh, surface wrapped and solved to convergence using a commercially available RANS solver (STARCCM+). The methodology employs representative boundary conditions, such as rotating wheels and a moving ground plane.A review is provided of the effect of cooling flows on the vehicle aerodynamics, compared to published data, which suggest cooling flow accounts for 26 drag counts (0.026 Cd). Further, a sensitivity analysis of the pressure drop curves used in the porous media model of the heat exchangers is made, allowing for an initial understanding of the effect on the overall aerodynamics.An analysis is made of the various pressure drops on the overall vehicle drag with a breakdown of the most affected components. Simulations suggest that at 50 kph, containing fan driven cooling flow, a linear relationship can be obtained where a variation of 1% in the pressure drop can result in the drag count changing by 0.02 drag counts (0.00002 Cd). An increase in the pressure drop is predicated increase heat exchangers drag, but a combination of the higher momentum loss in the heat exchangers and consequently lower flow velocities downstream at the engine and firewall, along with cooling outflow re-attaching to the under body of the vehicle combine for an overall predicted drag reduction.Further the system is studied without the radiator resulting in an increase of drag over the baseline case of 5 drag counts (0.048 Cd), indicating the need to reduce airflow downstream of the radiator. When blanking the radiator itself a drag difference of 18 drag counts (0.018 Cd) was obtained, further emphasizing the importance of the downstream region.