The ideal racing line through a turn will vary depending on a car’s traction limit, braking capability, and acceleration capability. The driver of a car with crappy acceleration may find that the fastest way around a track is to maximize the radius of his racing line through each corner; OTOH, when driving a car with excellent acceleration, it may be better to slow the vehicle much more to enable a high degrees-per-second turn rate, and then get on the gas afterwards to restore warp speed.
Given the amount of money involved in Formula One racing, I wondered about this. Do they use numerical simulations to determine the fastest way around any given track for any given car? I’m imagining a system where you could input:
-track geometry (turn radii, lane width, etc.)
-car characteristics (braking performance, lateral traction, power curve, etc.)
and the computer would produce an ideal racing line, including braking points, turn-in points, and so on, which could then be put into a real first-person driving simulator where the driver could practice following that line.
This seems like it might work well for finding the best solution without requiring hours and hours of expensive real-world track time, or wasting a driver’s time in a simulator by having him randomly try different things. OTOH, conditions in the real world vary considerably from moment to moment, so maybe a numerical model based on assumed/fixed values is too rigid, and a driver just has to see how the track and car are actually behaving on any given day, and adjust accordingly.
So anyway…do they, or don’t they, use numerical simulations to find the fastest path around a race course before ever putting a driver behind the wheel of a real car or a driving similator?
I learned from some tours around F1 constructors that they do everything you can think of and then some. They spend money like it is someone else’s - oh! it is someone else’s.
They have pretty accurate simulations for every track, based on the actual performance last time out. They know to several places of decimals, how much stress is placed on every single component in the car, every second of every circuit. They allow for rain, shine, hot, cold; they adjust for diminishing fuel load; for all I know they allow for the driver’s lost weight from sweat.
None of this does away with the need to spend the winter testing next year’s car to the limit and beyond.
I’m not a Formula 1 expert, but I am an amateur racer and an F1 fan. I’m not familiar with anything exactly like what you’re describing. Finding the academic “best line” based on geometry and car performance is a pretty basic activity, and not something that any halfway decent racing driver would need a computer to calculate for them.
At the level they’re running at, there’s no theoretical “best line” because it depends on frequently changing track and car conditions (temperature, fuel load, tire degradation) and driver preference. The entire skill of a racing driver at a competitive level is so much less about being able to execute the same line every lap, and so much more about being able to constantly adjust to changing conditions to extract that last .1%.
What they will do, however, is analyze the heck out of the data. There’s something in racing called the “ideal” or “perfect” lap, which is basically the lowest sum of all the driver’s best segments from practice. The team can help the driver approach this perfect lap by telling them which laps produced the best times in each segment, so the driver can the learn that “turn in at the patch of grass on turn 17” produces a lap time .1 seconds quicker than “turn in 3 feet before the patch of grass on turn 17.” Now the reason that first approach is slightly faster probably has less to do with basic geometry and more to do with the fact that, on that day, the outside of turn 17 has some rubber marbles built up, or there was a GT race the weekend before and there’s more rubber built up on the inside of the turn, etc.
They’ll also compare the segment times for each of their two drivers, hoping that one can learn from the other. Sometimes you can see them try this in practice and it works, and sometimes one driver just doesn’t execute the turn strategy the same as the other and it doesn’t pick up any time.