Wind and gradient at the Belgium GP
The F1 teams have returned to work after a much needed mid season break, so it is with great anticipation that we watch to see how the championship battle will unfold. With only 14 points separating Vettel and Hamilton even the smallest differences in car performance could swing the result.
The next chapter of the story unfolds at one of our favorites, Circuit de Spa-Francorchamps. Renowned for its large elevation changes and unpredictable weather, we asked ourselves how much difference does this make to car setup? Fortunately we are in the position to run quickly run 8000 DynamicLaps to find out, equivalent to changing car setup every single lap for 182 race distances.
In this simulation we have picked a downforce package for our car (see Aerodynamic Upgrades Facilitated: Isochronal Ratio for a clue of how you might do this) and we are looking at the effect of front and rear ride height. The ride height of the car plays a crucial role in getting the downforce and balance right through the corners. Front ride height is pretty uninteresting; a low front and lots of car rake generates more front downforce, but if you go too low you’ll wear the plank out and get disqualified. Rear ride height is a bit more subtle. The central green plot in Figure 1 shows our rear ride height curve; how CLiftTotalR (rear downforce coefficient, or CLR for short) changes with hRideR (rear ride height). At the highlighted data point we are at the peak of the rear ride height curve, with the car achieving maximum CLR of 2.23 at a rear ride height of 108mm. However if we move either side of this peak we’ll lose CLR.
Unfortunately we can’t just fix the rear at 108mm and move on, we need some suspension movement to help mechanical grip and traction, and even if we did have solid suspension there would be further movement that comes from tyre squash. In Figure 1, the car setup ride height (at 0kph) is 125mm, but the downforce generated by the car causes this to move down to the peak of 108mm at 162kph. Other factors to consider when setting hRideR are that drag (central red plot) increases as we raise hRideR, CLF (central orange plot) decreases as we drop hRideR, and of course we shouldn’t forget the effect it has on the car centre of gravity height. We then come on to the subject of balance; typically in low speed corners we have understeer, while in high speed corners bad things tend to happen if we don’t have enough CLR, so we can also use our setup rear ride height as a tool to affect our high speed to low speed aerobalance (blue plot: CLF/(CLF + CLR)).
It is no surprise that race engineers cannot work out in their head where to set the rear ride height for a specific circuit, particularly for races where there is mix of low speed and high speed corners to consider. Hence, all F1 teams run laptime simulations of varying degrees of quality to work out where to set the ride heights for a specific circuit. However, only a few F1 teams have the capability to add track gradient and wind to their laptime simulations so we wanted to see what effect this has.
We ran 2000 simulations on the baseline, then repeated with track gradient on, then added qualifying forecast wind (8mph WSW) and finally with race wind (5mph NNE). We turned up the plank coefficient of friction to penalise the laptime for runs with too much bottoming (alternatively we could have used some metric to match end of straight bottoming), and we didn’t change mechanical balance. The results are shown in the table below:
When we look into the results in Figure 2, perhaps the biggest surprise is that despite all the attention that is given to the elevation change at Spa, the net effect on laptime is only 20ms. When we look in more detail we lose 128ms through the first sector with 42m of altitude gained, but in sector 2 we gain 256ms back again with the 84m drop, leaving a 42m climb back to the start finish line where surprisingly we end up only 20ms behind. It will be interesting to see if this happens at other circuits on the calendar.
The effect of the forecast wind in qualifying, which is a side wind coming from the left on the start finish straight, makes a much bigger difference, gaining 490ms. This moves the optimum setup up +2mm at the front, and down -2mm at the rear. We didn’t lose much time after Eau Rouge, but gained a lot of time from the tailwind in the last sector.
However, the race wind is forecast to change direction and become a side/tailwind after Eau Rouge, which helps us gain 345ms down this straight but we lose some of this gain in the final sector, leaving this car 140ms quicker than no wind.
The real question is: if we didn’t adjust our setup to compensate for the wind direction, how much performance would we stand to lose? If we run all setups with the same ride height as the baseline we can see the results in Figure 3.
If we look carefully at the optimum setup from Figure 2 we can see that both the qualifying and race cars run better with a lower rear ride height, 122mm and 120mm respectively. While the qualifying car preferred a higher front ride height, it would be easy for the race engineer to achieve a similar balance effect by changing front wing angle which is allowed under parc fermé regulations. If we didn’t simulate the wind effect and ran all setups with the same ride height as shown in Figure 3, we would stand to lose up to 0.15sec in qualifying and 0.12sec in the race!
Interestingly the wind forecast for Sunday changed while I was writing this article. Fortunately using Canopy it only took minutes to re-run the relevant sims and update the article. Let’s hope the race enginner of <insert your favorite driver name> has such a powerful tool at their fingertips, because it could make all the difference in the championship battle. Enjoy the race!