- 331
- Addison, TX
- MX5_Racer_69
I don't know if this would be of any use to anyone. But here is an overview of the tuning method I use to come up with my tunes. You should be able to use this to come up with your own.
The method I've used to come up with my tunes is a variant of the computer programming algorithm called the BTree Search Algorithm. In a nutshell, if you have a list of items in sorted order, you can search that list for a specific value by consecutively dividing the list and comparing the value at that location to the value you are looking for. If the value at that location is greater than your search value, you split the difference backwards. If it's less, you split the difference forward. And so on. In the worst case, you find your search value in N/2 searches.
With tuning a car, we do basically the same thing. The difference is that we’re not searching a list, we are searching for a theoretical fastest lap time. And that means lots of testing.
We begin with what we know. We have a car that hasn’t been tuned. I will typically install all the tuning parts on the car before I begin but leave them with their default values to create a baseline lap time. Then I take the car out on the test track and start doing laps. I use the Cape Ring Inside track for my testing because it’s short and has several technical sections, allowing me to test the car and get a lap time in a short amount of time. After several laps, I end up with my baseline fastest lap. That’s our starting point.
The next step is to begin experimenting with different settings. Rather than just trying stuff a little at a time, I wanted a scientific way to get the optimum setting as fast as possible, and here is where we use our BTree Search Algorithm technique.
Lets say we have a slider for a tuning setting that goes from 0 to 100. And lets say the default for that value is 50. I will make a guess based on what I know about the settings and either go lower or higher. For example, if we’re talking about ride height, I’m going to go lower. How much lower should I go? No need to guess. The BTree Algorithm says divide in half. So we go midway between 0 and 50 and get a value of 25. Ok, how do I know if that’s a good value? We take the car on the track and do some laps. If the car is faster, we divide again in the same direction. If the car is slower, we can try between 50 and 100. Or, if we feel that the faster number is under 50, we can try somewhere in the 25 to 50 range. By dividing the interval in half each time and taking a measurement, we quickly (in about 3 or 4 tests) come up with our optimum value. This is the fastest way I’ve found to find the best settings based on actual lap times.
With this in mind, lets start tackling our biggest problems and work through them. A car from the dealer typically has several key problems. They are set from the factory with initial understeer for safety reasons. They have a very high center of gravity to handle pot holes and daily driving. And they are setup for long highway driving and optimized for gas mileage, not cornering. We need to fix all of that.
The first problem I tackle is center of gravity. It’s going to have the biggest impact on our cornering ability. So the question is how low we can go. Using the BTree method, I just start lowering the car until it stops getting faster. That’s step one.
Step two gets a little more tricky because ride height adjustments require spring adjustments to prevent the car from bottoming out. So to solve this problem, we have to begin BTree-ing the springs. Stiffen the springs ... then do some laps. If it needs to be stiffer, split the difference. If it needs to be softer, split the difference. At this point, there is some micro-tuning we need to do. Here is where I really learned a lot about tuning the car. As you are doing laps, keep a good eye on the tire temperatures. Ignore heating due to mistakes, but look at the tires on clean turns. If the front tires are heating up more than the rears, add some stiffness to the rear. And vice versa. Once you find the stiffness that gives the fastest lap times, hate to say it, but we're not done. It's kind of like tuning a guitar. Once you tighten one string, the other strings get a little looser. Same is true here. Now that we have tuned the springs, we have to see if changes to the ride height will again net faster lap times. So we make micro-adjustments at this point, keeping an eye on the tire temperatures as we did before.
Once we have ride height and springs, everything else is just small tweaks. You can do the BTree method if you like. Or you can just increment in fixed amounts and see what happens. Do this with toe, camber, sway bars, dampeners and finally, the LSD. After each adjustment, take the car out on the track. Are you faster with those changes or slower? How are the tires doing? Are they heating evenly or is there a bias front or back? Adjust and retest. I highly recommend using your lap replays to watch the tire temperatures. Helps a lot.
Once you’ve used this method, you end up with a tune that utilizes as much of the grip that you have in your tires during normal cornering. And this should, theoretically, make you faster.
That in a nutshell is how I come up with my tunes. Hope that helps.
The method I've used to come up with my tunes is a variant of the computer programming algorithm called the BTree Search Algorithm. In a nutshell, if you have a list of items in sorted order, you can search that list for a specific value by consecutively dividing the list and comparing the value at that location to the value you are looking for. If the value at that location is greater than your search value, you split the difference backwards. If it's less, you split the difference forward. And so on. In the worst case, you find your search value in N/2 searches.
With tuning a car, we do basically the same thing. The difference is that we’re not searching a list, we are searching for a theoretical fastest lap time. And that means lots of testing.
We begin with what we know. We have a car that hasn’t been tuned. I will typically install all the tuning parts on the car before I begin but leave them with their default values to create a baseline lap time. Then I take the car out on the test track and start doing laps. I use the Cape Ring Inside track for my testing because it’s short and has several technical sections, allowing me to test the car and get a lap time in a short amount of time. After several laps, I end up with my baseline fastest lap. That’s our starting point.
The next step is to begin experimenting with different settings. Rather than just trying stuff a little at a time, I wanted a scientific way to get the optimum setting as fast as possible, and here is where we use our BTree Search Algorithm technique.
Lets say we have a slider for a tuning setting that goes from 0 to 100. And lets say the default for that value is 50. I will make a guess based on what I know about the settings and either go lower or higher. For example, if we’re talking about ride height, I’m going to go lower. How much lower should I go? No need to guess. The BTree Algorithm says divide in half. So we go midway between 0 and 50 and get a value of 25. Ok, how do I know if that’s a good value? We take the car on the track and do some laps. If the car is faster, we divide again in the same direction. If the car is slower, we can try between 50 and 100. Or, if we feel that the faster number is under 50, we can try somewhere in the 25 to 50 range. By dividing the interval in half each time and taking a measurement, we quickly (in about 3 or 4 tests) come up with our optimum value. This is the fastest way I’ve found to find the best settings based on actual lap times.
With this in mind, lets start tackling our biggest problems and work through them. A car from the dealer typically has several key problems. They are set from the factory with initial understeer for safety reasons. They have a very high center of gravity to handle pot holes and daily driving. And they are setup for long highway driving and optimized for gas mileage, not cornering. We need to fix all of that.
The first problem I tackle is center of gravity. It’s going to have the biggest impact on our cornering ability. So the question is how low we can go. Using the BTree method, I just start lowering the car until it stops getting faster. That’s step one.
Step two gets a little more tricky because ride height adjustments require spring adjustments to prevent the car from bottoming out. So to solve this problem, we have to begin BTree-ing the springs. Stiffen the springs ... then do some laps. If it needs to be stiffer, split the difference. If it needs to be softer, split the difference. At this point, there is some micro-tuning we need to do. Here is where I really learned a lot about tuning the car. As you are doing laps, keep a good eye on the tire temperatures. Ignore heating due to mistakes, but look at the tires on clean turns. If the front tires are heating up more than the rears, add some stiffness to the rear. And vice versa. Once you find the stiffness that gives the fastest lap times, hate to say it, but we're not done. It's kind of like tuning a guitar. Once you tighten one string, the other strings get a little looser. Same is true here. Now that we have tuned the springs, we have to see if changes to the ride height will again net faster lap times. So we make micro-adjustments at this point, keeping an eye on the tire temperatures as we did before.
Once we have ride height and springs, everything else is just small tweaks. You can do the BTree method if you like. Or you can just increment in fixed amounts and see what happens. Do this with toe, camber, sway bars, dampeners and finally, the LSD. After each adjustment, take the car out on the track. Are you faster with those changes or slower? How are the tires doing? Are they heating evenly or is there a bias front or back? Adjust and retest. I highly recommend using your lap replays to watch the tire temperatures. Helps a lot.
Once you’ve used this method, you end up with a tune that utilizes as much of the grip that you have in your tires during normal cornering. And this should, theoretically, make you faster.
That in a nutshell is how I come up with my tunes. Hope that helps.
