For years, pressure data from the race-track and the wind tunnel, and simulated pressure in CFD for the floor leading edge did not match at all. Aerodynamicists tested all sorts of settings and pressure measurement methods but could not get to the bottom of this issue. Until, that is, particle-image velocimetry images tracking the Y250 vortex’s evolution in the wind-tunnel revealed the culprit.
Y250 vortex is the vortex shed by the front-wing inner tip and is critical in wheel-wake control and in driving flow into the floor leading edge.
The challenge in this search for missing floor-leading edge suction (between track-WT-CFD) was firstly to identify the correct problem to solve. That correct problem turned out to be finding the reason(s) for difference in characteristics (strength, extent, position) of Y250 vortex between track-WT-CFD.
WT PIV and floor pressure tap data from track when compared to CFD data, revealed that the Y250 stayed “stronger for longer” in the WT and on the track as compared to CFD (and hence all CFD simulations relied upon for designing aero-components downstream of the front-wing were incorrect by a long way). This also revealed that the phenomenon of sudden early dissipation of Y250 vortex in CFD, known as Vortex Bursting, could also happen in the WT and track at certain ride-heights and speeds.
Like a Matryoshka doll of problems, this then translated to the following –
To improve floor leading edge suction of the car on track-
1. Fix CFD to predict similar Y250 vortex performance (Track the vortex core in CFD (using a bespoke core tracking tool) and WT (using PIV) to then tweak CFD settings for better correlation)
2. Make Y250 stronger and prevent bursting
For challenge #1 of fixing CFD vortex performance we reached out to academia and turbulence model experts who had a solution known as Curvature-Correction (CC). Introducing CC with a K-w turbulence model in Fluent and in Star CCM+, reduced vortex dissipation significantly. This meant CFD Y250 vortex was now nearly as ‘strong and long’ as the wind-tunnel one at all ride conditions. This was verified using PIV data over a range of wind-tunnel tests.
Then, challenge #2 was tackled by improving front-wing design to make the Y250 vortex stronger at source. It also involved redesigning mid-car devices like the Bargeboard, to reduce adverse pressure gradient encountered by the Y250 vortex. (Adverse pressure gradient can induce bursting by changing the ratio of axial to tangential velocity of a vortex)
All this took months, approximately 9-12 months, but over this period gradual improvement in underfloor suction was successfully achieved.
Improved correlation between the track, wind-tunnel and CFD also meant that car performance improved at a faster rate than it did before i.e. more parts worked as expected, on the track.
A key lesson from this project was to tackle complex challenges in tandem and in small iterations. Large changes often lead to no solution when solving complex non-linear problems.
