By Ales Alajbegovic, PHD, Vice President, Ground Transportation Applications
We’ve all experienced turbulence on an airplane, but back on terra firma, those troublesome, swirling pockets of air can have just as great an impact on how a car moves along the road and how much fuel it spends. And for automakers, the ability to accurately simulate turbulence helps to create quieter, slipperier, and more fuel-efficient vehicles.
The airflow around a car is always turbulent, particularly in the critical boundary layer – the few millimeters closest to the vehicle’s surface – and in the wake, the vehicle leaves behind it. Turbulent flow is also transient, as opposed to steady-state: it continuously changes with time – smoke wands will show this in a wind tunnel, and the phenomenon is reflected in the measurements that aerodynamicists record during tunnel testing. In addition, the air on the road is never steady. There are traffic turbulence and wind affecting the real world aerodynamic performance of the vehicle.
Logically, any digital simulation of airflow should take account its transient nature to accurately replicate the real-world behavior. If you ignore the transient nature of the airflow, it is impossible to predict the exact aerodynamic force on the car.
Simulation software needs to calculate how turbulence interacts with the car and how it affects pressure distribution around it. It can also take into account the real world effects like traffic turbulence and wind. Vehicle wake is always transient, and its correct prediction requires the ability to simulate transient flows. This is not possible with steady-state codes. Exa technology is underpinned by the Lattice Boltzmann Method (LBM) algorithm, the most efficient and comprehensive numerical scheme for simulating transient flows.
Accurately predicting aerodynamic forces is crucial to delivering the low drag coefficients that reduce fuel consumption. By modeling how turbulence changes in time, simulation software based on transient airflow can correctly predict flow structures and provide simulation accuracy reaching±1 count (equivalent to a CD of 0.001), as opposed to an estimated ±30 counts for the steady-state codes, which equates to an approximate 5% error in the predicted fuel economy.
Our supporting work with Tesla on the Model S is a good example. Electric vehicles present unique aerodynamic challenges to engineers as the flat underbody is prone to creating unstable wakes that can make the vehicle ride potentially unsafe. In addition, efficient aerodynamics is very important for helping reduce drag and increasing vehicle range. Using steady-state flow simulations, Tesla aerodynamicists simply couldn’t reach a steady state’, with the drag fluctuating significantly and unable to settle at a converged value. This was when they turned to Exafor help, which we did. Simulations with our transient flow simulator showed strong sidewise oscillations of the wake – something also observed in the measurements. Once these oscillations were identified and understood by analyzing the simulation results, it was possible to design the rear diffuser you see on the Model S today which is there to stabilize the vehicle wake. This also helped lower vehicle drag.
Another common example is where design teams have a specific design intent like a rounded C-pillar which will usually cause unstable wakes behind cars. In such cases special effort is needed to stabilize the wake. Engineers can employ devices that initiate a separation to control the wake – be it a small edge or a spoiler. Transient turbulent flow down-stream from these devices needs to be investigated in order to predict how these devices are going to work – and this is only possible by studying transient flow behavior. Steady-state codes cannot find a ‘steady state’ because it doesn’t exist – you’ve got something, but it doesn’t help you.
For this reason, the real competition to transient flow simulation are only wind tunnels, and even they cannot accurately predict vehicle aerodynamics on the road. Because of their limited size the flow around a car is never the same as on the road. They also cannot reproduce on-road effects like wind or traffic turbulence. These are critical for the real-world fuel economy predictions that will be mandated by regulations and demanded by the public. Transient flow simulation can accommodate all these issues, and it will be the critical tool to help the automakers meet both regulations and public expectations.