I worked up a spreadsheet to model the tradeoff between turning performance and braking performance.

The goal was to be able to enter an arbitrary target braking g, and then have the spreadsheet compute the effect on lateral g, and ultimately lap time. I didn’t get that far (I’d need to develop a state machine, and excel just isn’t good for that), but I was able to run the calculation for a single, theoretical, compromise between braking and turning. I chose a target braking acceleration of -1g, and figured that at that camber setting, my car would be capable of 1.04g of lateral acceleration. (This seems very reasonable given the figures I calculated from the back of Car and Driver which showed that most performance cars leave the factory with slightly more braking acceleration than turning acceleration)

I dumped a file of my fastest lap, with a row every 10th of a second and columns for time, distance, speed, lat acc, and longitudinal acc.

I then labeled each row as ‘Accelerating’ if the speed was up from the previous row. If not accelerating, I compared the lat and long acc, and chose ‘Decelerating’ if the magnitude of the longitudinal accel was greater and ‘Turning’ if the lateral accel was greater.

One very important thing to note is that only 13% of the lap (by time) is spent braking.

Then, in each of the ‘turns’, I computed a new speed based on the square root of the ratio of new lateral acceleration to original lateral acceleration. In other words, I slowed the turns by about 6% to account for the 12% drop in lateral acceleration g’s.

Working forward from the exit speed of each turn, I set the acceleration equal to the acceleration from the original data. That is, if the original data were accelerating by .5mph per .1 second, then that’s the way it worked for the new calculated speed—if you exited the turn 2mph slower, you stayed at exactly 2mph slower for the ensuing straight. I made exceptions for the front and back straights, where aerodynamics and terminal velocity start to come into play. In these cases, I looked up the acceleration for the computed speed based on the acceleration seen in the original data when the speeds were identical. That is, if I’m going 59 mph, my speed on the next row will be increased by the same amount as when I was going 59 mph from the original data. What’s interesting about this is that if you leave Oak Tree 2.2 mph slower, you are still going 1.7 mph slower when you hit the end of the straightaway.

Finally, I worked backwards from the turns in order to calculate the effect of the improved braking performance. I kept going back until I was toward the beginning of the braking zone. I then iterated forward on acceleration and back on braking until the two speeds matched. (Any working forward of the accelerating meant delaying the brakes—this is where the improved braking performance would pay off).

I then calculated the average speed of the lap, by averaging the speed column. This is easily converts to elapsed time, and a time difference is calculated.

So, the answer is: roughly equivalent braking and turning results in a lap time that is 5 seconds slower.

Check the Speed vs Time tab, and this graph tells the story.

First, the real data show me taking the climbing esses at ~100mph (when my setup had max lateral of 1.18g). A six percent drop means that with only 1.04g to work with, I’d be running at 94mph. And this story plays out, essentially, on 87% of the lap!

The payback, under braking—well here’s the rub: since you end up needing to slow down more (apex speed of corner 1 is 2 mph slower) you really don’t get to shorten the braking zone very much.

Now, there is a major methodological problem here. Since everything is reported by time, I’m not accounting for the fact that if you’re going slower down a straight, it takes longer, and thus you have more time to accumulate more speed. So, I need to start from scratch on this spreadsheet (or find a better way of simulating), where each row is reported by distance (or redo all my acceleration and decelerations to account for the fact that going slower gives you more time to gain (or lose) speed). But, this was a decent first order approximation, and I doubt that it would make much difference.

Now, there is a major methodological problem here. Since everything is reported by time, I’m not accounting for the fact that if you’re going slower down a straight, it takes longer, and thus you have more time to accumulate more speed. So, I need to start from scratch on this spreadsheet (or find a better way of simulating), where each row is reported by distance (or redo all my acceleration and decelerations to account for the fact that going slower gives you more time to gain (or lose) speed). But, this was a decent first order approximation, and I doubt that it would make much difference.