First, sincere apologies to faithful readers for the long, long dry spell in new blog posts. I’ve been busy with two racing programs. That’s the reason, but it’s no excuse. Oh, well.
If you followed the 2010 ALMS race weekend at Lime Rock, then the topic of this blog will come as no surprise at all. Let’s end the dry spell in blog posts by talking about rain.
I’ll be the first to tell you that I am never totally comfortable with rain setups, despite having run in the rain quite a bit and even having engineered a major series race win in the rain. Rain setups are always a conundrum. First, you don’t get a lot of opportunities to develop your rain setup. Then, there’s the really big question of how far do you take the rain setup. Then, there’s the other really big question of how skilled and comfortable your driver is on a wet track. Compared to those, the technical issues of the rain setup are almost “cut and dry”, to make a weak pun. Contrary to what the advice columns say, I’ll tackle the easy part first.
For starters, let’s assume we have a properly wet track (at least light rooster tails), with rain still falling at a steady enough rate to keep the track from drying. It’s been raining for a while, and there isn’t much heat left in the pavement. We don’t anticipate that it will quit raining during the session.
Let’s start with a small handful of basic principals. We need grip. The brake bias needs to be shifted to the rear. We absolutely cannot tolerate understeer. Let me say that one more time. We absolutely cannot tolerate understeer.
Chasing grip first…
First, we need all the downforce we can produce. Wing(s) to the maximum, more rake, add on all the gadgets, do it all. Drag be damned. That much, I know for sure. It’s essential that we remain respectful of the car’s fundamental aero characteristics. Pitch and ride height sensitivity will affect decisions to adjust rake, as well as spring or bump rubber rate decisions. I believe that as you add total downforce, you need to try to keep the aero balance as far forward as possible. All in the interest of avoiding understeer.
In pursuit of grip, it’s tempting to dramatically open the bump and rebound bleeds in the shocks. On the other hand, we need to remain respectful of the car’s needs for aero platform control, which may or may not change in the rain. We also have to remember that the bumps, curbs, and other surface irregularities in the track will not go away just because they get wet. So, how much bleed is enough? It depends.
In race series where the rules allow driver-adjustable anti-roll bars, it’s easy to soften the bars for some grip. Perhaps equally important, it’s easy to reverse the adjustment. In other series, it becomes more of a judgment call whether to soften the bars, and if so, how much.
So that’s it for the easy changes. Now, if it’s really clear that the rain will continue, and at a track-soaking pace, we can also consider softening springs and reducing negative camber. Again, we need to be cognizant of platform control, if we choose to soften springs. It’s much safer to soften dramatically on a low-dowforce car than on a high-downforce one. Meanwhile, the camber change can be quite helpful, presuming the weather has the courtesy to stay properly wet. If not, you’re in trouble, aren’t you?
Finally, let’s have a look at the rain tires themselves. How fresh are they? If they’ve been previously run on a drying track, they will have worn away the chiseled edges that are so essential to cutting through the water to the track. If it’s really wet, fresh rains always help.
Now, a brief word about brake bias…
Braking Gs in the rain won’t be as high as in the dry, so we won’t have as much forward weight transfer. Lacking that, we’ll need less front brake bias. It’s pretty easy to make a rough calculation of the needed change. The whole process is almost painless if you have driver-adjustable bias and brake pressure sensors wired into onboard math in the data logger and displayed on the dash. A word to the wise – brake bias adjusters can be cranky little pieces of crap, and the time to debug them is not when the rain starts falling.
OK, so let’s talk about understeer…
Understeer in the rain is more than just an inconvenience. It’s an opportunity to arrive at a corner and plow straight off into the muddy grass, gravel, wall, or whatever other dramatic and damaging fate awaits the car. Not good, and it won’t endear you to either the driver or the mechanics.
Let’s remember that our dry-track handling balance assumes a lot of weight transfer to the front in braking, helping corner entry. Then, on exit, the rear lateral grip is reduced with an aggressive application of power. Neither condition is there for us in the wet. It will be necessary to shift the handling balance in the oversteer direction. It’s not as simple as aggressively softening the front springs and bar, because that might lose aero or roll platform control in a way we won’t like. You should work within the tuning options you’ve proven for your car in the dry.
The oft-mentioned option to disconnect both the front and rear anti-roll bars can, in the right circumstances, be a good move for preventing understeer. The wheel rate of the front bar is almost always higher than the rear. Nuff said.
Here are two serious “gotchas” related to cooling…
First, be careful of your engine temperatures. Some ECU calibrations may default into “cold-start” mode if temperatures drop too much.
And, some brake materials don’t take kindly at all to getting wet or getting too cool.
Everyone wants to just lay a strip or two of duct tape over the radiator opening or the brake ducts. Step back for a minute and think about how well duct tape will stick to a wet, dirty, oily surface, and you will see why I strongly prefer sheet aluminum blanking plates held in place by Camlocs.
Now, let’s talk drivers for a minute…
Some are really good in the rain, some are OK, and some are, well, not OK. I hope yours is in the first group. If so, you’ll look pretty good, no matter how well the car works. I once had a special case of a driver who was good in the rain, but didn’t care for losing any of his cherished platform support to softer springs, bars, or shocks. We just turned up the wing, reset the brake bias, bolted on the rains, and went off to win the race. As always, tune to the driver’s comments.
So, here’s the ugly question…
How far do you go with “messing up” your nice dry setup for the rain?
Here are some of the questions we’ll need to consider:
Will the rain let up soon, or get worse?
Is it real rain, or just a shower?
How much of the track is wet?
What can we learn from radar and the sky?
When will the track be wet enough to need rains, or dry enough to take them off?
How much heat remains in the track surface?
If this is a practice session, do we expect to qualify or race in the rain?
Do we even go out on the track?
All questions are for another day. Hopefully soon. See you later.
Tuesday, September 7, 2010
Wednesday, February 17, 2010
Finding the Setup, Part 2
Have a look at this Wikipedia entry. While not written specifically about racing, it relates directly to Jeff's "secondary effects".
http://en.wikipedia.org/wiki/Law_of_Unintended_Consequences
In response to a request on the FSAE forum, Jeff has kindly taken the time to make a PDF of his setup change spreadsheet. Note that all the second page shows the selections for the pull-downs (not seen) that are used to fill in the individual setup changes on the sheet.
Jeff Braun Setup Change Database PDF
http://en.wikipedia.org/wiki/Law_of_Unintended_Consequences
In response to a request on the FSAE forum, Jeff has kindly taken the time to make a PDF of his setup change spreadsheet. Note that all the second page shows the selections for the pull-downs (not seen) that are used to fill in the individual setup changes on the sheet.
Jeff Braun Setup Change Database PDF
Saturday, February 13, 2010
Finding the Setup
Here's another guest post, this time from Jeff Braun. Jeff's a successful working race engineer, concentrating these days on endurance sports car racing, but with a broad background in many racing series.
He also has the dubious distinction of being the only race engineer who actually drove against me. It was a while ago... If I recall, I beat him then, but he's paid it back as an engineer, more than once.
Since we're friends, I'll indulge him for scooping me on material that was to have been included in the "Philosophy" series. But, there's much more than that in his post. I'm sure you'll find it thought-provoking.
The Setup – How to get it right
Apart from money, a good setup may be the most sought after thing in racing.
One thing before we talk about how to get the setup right. Who is ultimately responsible for the setup on the car? The driver! Not the engineers, not the team manager, not the chief mechanic, or anyone else except the driver. I hate it when people say “the engineer really has that car set up well.” It has little to do with the engineer. I have yet to see an engineer put a setup on a car, install a new driver, never change the setup, and win everything in sight. It never happens. The driver guides the setup process, he directs the engineers in the areas that need improvement, he is in charge of the car, and he has to get it to his liking to be fast. The engineer only interprets what the driver tells him and comes up with suggestions on what changes could help the problems the driver has identified as the main thing preventing him from going quicker. The engineer has the technical resources to come up with the best change to get the desired results, but he can’t identify and prioritize the areas that need improving without a technically good driver to point him in the right direction.
The magic setup – No numbers
Getting the setup right is not about cambers, toe, springs, and all the technical numbers. It is about the approach you take and the procedures you use to get there. So, I am not going to tell you the perfect springs to run at Laguna, or what wing setting works best at Montreal. Why? Because, it all depends on way too many things. I just plain don’t know the answer. What I am sure of is that, if you follow the things below, you will be ahead of the game in finding the springs or wing settings for your driver, your car, and the track on the weekend you are racing.
Best way to a great setup – Don’t change anything!
So, you spend hours with your driver and the data in the shop designing the starting setup for the weekend. You have experience with the car and kind of know the type of track you are going to, but have never been there before. You have it nailed, you think. You unload and head out for the first session. Four laps later, the driver is in the pits looking for a change. You jump in and start changing things; you want it to be better for him.
After the session you make more changes, maybe three or four, because he was 1 second off the pace. Next session the car is doing different things than before. Oh no, the changes were in the wrong direction. You remove the three changes and go the other way with four more. Now it’s better, but now a there is a different problem.
The data looks confusing, the driver is confused, but time is running out now, and you are still 0.8 seconds off. Something must be way different about this track than you thought. You decide you may have to rethink the entire setup, or that you may not really understand the car like you thought.
The next session you are better. The times are closer to fast guys and the driver just needs a few small changes. The changes help and the panic subsides some. But, you still need three tenths.
Now, it’s qualifying. The changes helped, and you’re now two tenths off the pole and on the second row. So, let’s look at the qualifying setup. It is exactly how we unloaded! We went in a big circle.
I see this all the time, at all levels of motorsport. I don’t know of any race engineers who get paid by the change. Slow down. Let the driver learn the track, and then figure out what he needs. The track will clean up and get some rubber down. Have confidence in your starting setup. Don’t be in a hurry to change it all around right away. Talk with your driver and write out a testing plan listing each session and what you hope to get done. This will only be a guide for the weekend, but it allows you to plan the sessions as you wish they would go.
A driver’s feed back is junk for the first session, anyway. He is getting a feel for the track, the grip level, and the car. Use that session to set the ride height (as the best engineer/driver I have ever known told me - lower is better always, always), log some baseline data, check the car, read the tires, and get some fuel consumption data. Robbie Groff used to tell me he needed the car to “talk to him” before he could tell me what we needed. Don’t screw up a good setup in the first session. Think where our driver could have qualified if he had kept the setup he ended up with and progressed forward in the sessions instead of going in a circle.
The spec car
If you have a restricted car or spec car, you have less things to try on the weekend. You should know the car well, if you have some experience with it. Otherwise, you can get a good starting set up from someone in the series. These cars are really relatively simple and require a methodical approach, rather than the shotgun approach of multiple wild guesses in search of the magic set up that gives you the 1 second you need. Now, there are some spec cars that are so bad that you do need the tricks that only the veterans know. My advice is to be prepared for a frustrating time, buy or steal the tricks, or run a real race car series.
Most series in North America have good cars that behave as expected. Try to stay within your setup window. You must define that window and keep refining it as you get to know the car better. But, resist the urge to go way out of what you know on a race weekend. Save that for testing. Keep a list of things you wanted to try at a race weekend, but did not because you did not know what it would do. Answer those questions at the next test.
On race weekends, it is always best to make changes that produce known results. If you don’t know what a change will do, go testing. If things are going well and you can afford to test for a session, then give some of the unknowns a poke. You can always go back to your baseline. Just remember, the other drivers will be moving forward in their setup when you are testing. You may find yourself behind by losing the session.
The track will tip your base setup some, but not much in a spec car. Try to come up with a base setup which will be refined each weekend. Run it at each track for the first session or two. You know how the car performs and feels with this setup, you know you liked it at the last race and there are no surprises to it. If the car feels different at this new track, you have just quantified the track. Using your base setup for the car, you can see what the track did to it. Make adjustments in the window of what you know works. Be sure the car is talking to you before you start. My driver/engineer friend likes to say “the laws of physics don’t change when you cross state lines.”
One last thing on this subject. Don’t fall into what I call the “Runoffs mistake”. A driver does a great job all year and qualifies for the prestigious SCCA Runoffs by hard work, good understanding of the car, and good procedures to get the set up right. He makes the trip to Road America and knows he has a shot to win this thing. But he knows there are some very good drivers from all over the U.S., and he thinks he needs something special to win. He finds a trick tire that he has never run but it is “worth 0.3 seconds”. The expert shock guy gave him the best Road America set up. The engine guy has the new development exhaust system. The result – the poor guy gets clobbered, not only by the out of division drivers, but also by the guys he beat all year, who now run away from him. Run what you know and tune it better than the next guy. Save the tricks for winter testing.
Changes – The secondary effects
In talking to engineers and drivers, I hear that they made a change and it did not help or did something completely different than they expected. Most of the time, the reason is what I call the secondary effect of a change.
Most racers have a good idea of what a change will do to the car, or at least they know what they want it to do. The problem is that seldom does a change only do one thing to the performance of the car. Each change has a primary effect and a secondary effect. There are third and fourth effects I am sure, but I am just barely smart enough to figure out the second effect.
If you are trying to decide between a few changes to improve something in the balance of the car, make a list of the three or four possible changes and what you expect each to do. For example, assume we have an understeer from the apex to the exit. We could change:
1 – More front wing
2 – More rake
3 – Stiffer rear springs
4 – More front rebound
There may be fifteen other things you could do, but list the top contenders. Now, list what you think the secondary effect of each might be:
1 – More front wing – oversteer in high speed turns
2 – More rake – nervous rear in high speed turns
3 – Stiffer rear springs – worse power down traction
4 – More front rebound – harsher in the bumps
Look at the list and try to find a change that has a primary effect and secondary effect that are in the same direction. Also, consider what other problems the driver may be having, which we’ll call the secondary complaint. In our example, if the high speed turns were good or bordering on oversteer, then we can eliminate the wing and rake. The stiffer rear springs will keep the dynamic rake in the car (primary effect) and free the car up on exit, while with less power down traction (secondary effect). Everything works in the direction we want.
What we want to avoid is a change with the primary effect opposite from the secondary effect. Often, this results in the driver comment of “can’t really tell much difference.” When a change has an unexpected result, you should look at the secondary effect. Sometimes, what you thought was the secondary effect was really the primary effect, and it was in the opposite direction.
Keeping track of all these effects and changes can get very confusing. The problem is that the effect of a change on a FF-1600 is very likely to be different from the same change on a Formula Atlantic car.
The change data base
There is a way to help sort the changes and what they did to your car. It helps you learn the car and can be a great tool to suggest changes that you know the result of like we talked about earlier.
One year at the test days for the Daytona 24 Hour, I was done with a session and walked past the factory Nissan NISMO team from Japan. In their pit were four racks of computers and more electronics than a modern day F1 team has. I watched for a while, as the Group C Nissan went around and around. There were only two guys manning the computer banks, so I walked up and asked what they were logging. The answer was that they were logging over 200 channels of data. I tried to think of 200 channels I would want to have logged. I got to about 30 and had to ask “what are you guys logging?” The Japanese engineer said “mostly engine parameters”. He said the engine block had twenty strain gauges cast into the block to measure stress in various sections of the block.
I thought that was very cool, but wanted to know what they were going to do with the data. The engineer said that, sometime in the future, a Nissan engineer was going to want to know what the stress was in an engine block. When that happened, the data would be there. He said that the cost of running a race car was so high, that collecting data while it was running reduced the cost of gaining that information later, when someone needed it. That impressed me with the long term thinking to collect data. Collect the data when you can, sort it out when you have time.
Any race team can do the same thing, on a smaller scale. Using a spread sheet, make a matrix with each column being a parameter in your setup. Include columns for driver comment, change, and result of the change. Each row will be a track outing. Start to log the result of each change you make to your car. In short order, you will have a history of each change made to your car.
Using the data sort function of the spread sheet, you can sort the data. For example, you can show every time you had exit understeer. Then, look at what you changed when you had that problem and what the result was. This will prevent you from making the same set up mistake twice. And, it will tell you what things have worked to correct an exit understeer in the past. It will keep you in the window of your knowledge. When testing, look at the data base and try some things that you have not tried before. This expands your understanding of the car and makes your data base more helpful at the track.
Using a data base program, like Access, takes it to the next level with tables, forms, and special queries to make the search results more meaningful and detailed. If you spend some time (it took me about 30 hours and I am NOT an Access expert), you can answer more questions. You can ask the database to show what we changed to reduce a mid corner oversteer in a third gear turn when we had tire pressures above 15 psi and a rear wing setting below 10 degrees.
Remember, the sooner you start collecting the data, the cheaper every mile you do becomes, because you are getting more value for your dollar. I can say that this has helped me a bunch over the years. My Daytona Prototype data base has over 900 changes logged in it. It just makes me a better guesser.
So, to recap…
1 – The driver is in charge of his setup. The engineer just makes the change he thinks will help him the most.
2 – Resist making changes until the driver really has a good feel for the car. It has to “talk to him” first.
3 – Physics does not change when you cross state lines. Go with what you know.
4 – The secondary effect of a change can be in the opposite of the primary effect. Never make a change without considering the secondary effect.
5 – Use a data base to increase your understanding of your car quicker and with more accuracy.
6 – If a change does not have the effect you thought it would, than you are missing some effect that you did not consider. There is not some weird phenomenon going on, you just don’t understand the circumstances of the particular situation.
7 – When in doubt, go back to your base setup and start over from there.
8 – Never copy another faster team’s setup. You need to know why yours does not work, so you can be better next time.
9 - If a change works the way you thought it would, you did not learn anything. You did become faster, which is always a good thing. But, when the change does not work as planned, you have a great chance to become smarter. Grab on to that and figure it out.
10 – When recording changes in your notes, write down why you made that change, your thinking on what you expect it to do, and why. Then, you can go back later and see what your thinking was for making that change and decide where your thinking was wrong. This may happen months later, as you get to know the car better, but it allows you to see where your mistake was, not just that it was a mistake.
OK, one more thing.
My friend, the driver/engineer mentor of mine, wrote these points down once about problem solving. Think about them. It may change your approach to finding the perfect set up.
1 – What’s right is right and everything else is wrong to some degree.
2 – What is REALLY happening here?
3 – Nothing happens for no reason.
4 – Everything is attributable.
5 – If X is true then Y must also be true. If I can’t prove that Y actually does as I predict, then I don’t know anything at all about X.
6 – What I am certain is correct can change instantly in the light of what is REALLY correct, whether I like it or not.
7 – Just because I don’t want to believe it, doesn’t make it wrong.
7a – Just because I want to believe it doesn’t make it right, either.
8 – Knowing what is wrong is every bit as important as knowing what is right.
9 – If it isn’t all the things you think it is, then it is something else. (Sherlock Holmes)
10 – You only know something if you can prove it. Everything else is “I suspect” or “I guess” or “I wonder if” or “it is my theory that...”
11 – The right answer is still the right answer even if you didn’t think of it.
12 – The right answer is still the right answer even if you don’t have any idea of why it works... but find out later for sure, because the underlying principles will always apply.
13 – Asking other people for answers is perfectly acceptable, as long as you never believe them.
He also has the dubious distinction of being the only race engineer who actually drove against me. It was a while ago... If I recall, I beat him then, but he's paid it back as an engineer, more than once.
Since we're friends, I'll indulge him for scooping me on material that was to have been included in the "Philosophy" series. But, there's much more than that in his post. I'm sure you'll find it thought-provoking.
The Setup – How to get it right
Apart from money, a good setup may be the most sought after thing in racing.
One thing before we talk about how to get the setup right. Who is ultimately responsible for the setup on the car? The driver! Not the engineers, not the team manager, not the chief mechanic, or anyone else except the driver. I hate it when people say “the engineer really has that car set up well.” It has little to do with the engineer. I have yet to see an engineer put a setup on a car, install a new driver, never change the setup, and win everything in sight. It never happens. The driver guides the setup process, he directs the engineers in the areas that need improvement, he is in charge of the car, and he has to get it to his liking to be fast. The engineer only interprets what the driver tells him and comes up with suggestions on what changes could help the problems the driver has identified as the main thing preventing him from going quicker. The engineer has the technical resources to come up with the best change to get the desired results, but he can’t identify and prioritize the areas that need improving without a technically good driver to point him in the right direction.
The magic setup – No numbers
Getting the setup right is not about cambers, toe, springs, and all the technical numbers. It is about the approach you take and the procedures you use to get there. So, I am not going to tell you the perfect springs to run at Laguna, or what wing setting works best at Montreal. Why? Because, it all depends on way too many things. I just plain don’t know the answer. What I am sure of is that, if you follow the things below, you will be ahead of the game in finding the springs or wing settings for your driver, your car, and the track on the weekend you are racing.
Best way to a great setup – Don’t change anything!
So, you spend hours with your driver and the data in the shop designing the starting setup for the weekend. You have experience with the car and kind of know the type of track you are going to, but have never been there before. You have it nailed, you think. You unload and head out for the first session. Four laps later, the driver is in the pits looking for a change. You jump in and start changing things; you want it to be better for him.
After the session you make more changes, maybe three or four, because he was 1 second off the pace. Next session the car is doing different things than before. Oh no, the changes were in the wrong direction. You remove the three changes and go the other way with four more. Now it’s better, but now a there is a different problem.
The data looks confusing, the driver is confused, but time is running out now, and you are still 0.8 seconds off. Something must be way different about this track than you thought. You decide you may have to rethink the entire setup, or that you may not really understand the car like you thought.
The next session you are better. The times are closer to fast guys and the driver just needs a few small changes. The changes help and the panic subsides some. But, you still need three tenths.
Now, it’s qualifying. The changes helped, and you’re now two tenths off the pole and on the second row. So, let’s look at the qualifying setup. It is exactly how we unloaded! We went in a big circle.
I see this all the time, at all levels of motorsport. I don’t know of any race engineers who get paid by the change. Slow down. Let the driver learn the track, and then figure out what he needs. The track will clean up and get some rubber down. Have confidence in your starting setup. Don’t be in a hurry to change it all around right away. Talk with your driver and write out a testing plan listing each session and what you hope to get done. This will only be a guide for the weekend, but it allows you to plan the sessions as you wish they would go.
A driver’s feed back is junk for the first session, anyway. He is getting a feel for the track, the grip level, and the car. Use that session to set the ride height (as the best engineer/driver I have ever known told me - lower is better always, always), log some baseline data, check the car, read the tires, and get some fuel consumption data. Robbie Groff used to tell me he needed the car to “talk to him” before he could tell me what we needed. Don’t screw up a good setup in the first session. Think where our driver could have qualified if he had kept the setup he ended up with and progressed forward in the sessions instead of going in a circle.
The spec car
If you have a restricted car or spec car, you have less things to try on the weekend. You should know the car well, if you have some experience with it. Otherwise, you can get a good starting set up from someone in the series. These cars are really relatively simple and require a methodical approach, rather than the shotgun approach of multiple wild guesses in search of the magic set up that gives you the 1 second you need. Now, there are some spec cars that are so bad that you do need the tricks that only the veterans know. My advice is to be prepared for a frustrating time, buy or steal the tricks, or run a real race car series.
Most series in North America have good cars that behave as expected. Try to stay within your setup window. You must define that window and keep refining it as you get to know the car better. But, resist the urge to go way out of what you know on a race weekend. Save that for testing. Keep a list of things you wanted to try at a race weekend, but did not because you did not know what it would do. Answer those questions at the next test.
On race weekends, it is always best to make changes that produce known results. If you don’t know what a change will do, go testing. If things are going well and you can afford to test for a session, then give some of the unknowns a poke. You can always go back to your baseline. Just remember, the other drivers will be moving forward in their setup when you are testing. You may find yourself behind by losing the session.
The track will tip your base setup some, but not much in a spec car. Try to come up with a base setup which will be refined each weekend. Run it at each track for the first session or two. You know how the car performs and feels with this setup, you know you liked it at the last race and there are no surprises to it. If the car feels different at this new track, you have just quantified the track. Using your base setup for the car, you can see what the track did to it. Make adjustments in the window of what you know works. Be sure the car is talking to you before you start. My driver/engineer friend likes to say “the laws of physics don’t change when you cross state lines.”
One last thing on this subject. Don’t fall into what I call the “Runoffs mistake”. A driver does a great job all year and qualifies for the prestigious SCCA Runoffs by hard work, good understanding of the car, and good procedures to get the set up right. He makes the trip to Road America and knows he has a shot to win this thing. But he knows there are some very good drivers from all over the U.S., and he thinks he needs something special to win. He finds a trick tire that he has never run but it is “worth 0.3 seconds”. The expert shock guy gave him the best Road America set up. The engine guy has the new development exhaust system. The result – the poor guy gets clobbered, not only by the out of division drivers, but also by the guys he beat all year, who now run away from him. Run what you know and tune it better than the next guy. Save the tricks for winter testing.
Changes – The secondary effects
In talking to engineers and drivers, I hear that they made a change and it did not help or did something completely different than they expected. Most of the time, the reason is what I call the secondary effect of a change.
Most racers have a good idea of what a change will do to the car, or at least they know what they want it to do. The problem is that seldom does a change only do one thing to the performance of the car. Each change has a primary effect and a secondary effect. There are third and fourth effects I am sure, but I am just barely smart enough to figure out the second effect.
If you are trying to decide between a few changes to improve something in the balance of the car, make a list of the three or four possible changes and what you expect each to do. For example, assume we have an understeer from the apex to the exit. We could change:
1 – More front wing
2 – More rake
3 – Stiffer rear springs
4 – More front rebound
There may be fifteen other things you could do, but list the top contenders. Now, list what you think the secondary effect of each might be:
1 – More front wing – oversteer in high speed turns
2 – More rake – nervous rear in high speed turns
3 – Stiffer rear springs – worse power down traction
4 – More front rebound – harsher in the bumps
Look at the list and try to find a change that has a primary effect and secondary effect that are in the same direction. Also, consider what other problems the driver may be having, which we’ll call the secondary complaint. In our example, if the high speed turns were good or bordering on oversteer, then we can eliminate the wing and rake. The stiffer rear springs will keep the dynamic rake in the car (primary effect) and free the car up on exit, while with less power down traction (secondary effect). Everything works in the direction we want.
What we want to avoid is a change with the primary effect opposite from the secondary effect. Often, this results in the driver comment of “can’t really tell much difference.” When a change has an unexpected result, you should look at the secondary effect. Sometimes, what you thought was the secondary effect was really the primary effect, and it was in the opposite direction.
Keeping track of all these effects and changes can get very confusing. The problem is that the effect of a change on a FF-1600 is very likely to be different from the same change on a Formula Atlantic car.
The change data base
There is a way to help sort the changes and what they did to your car. It helps you learn the car and can be a great tool to suggest changes that you know the result of like we talked about earlier.
One year at the test days for the Daytona 24 Hour, I was done with a session and walked past the factory Nissan NISMO team from Japan. In their pit were four racks of computers and more electronics than a modern day F1 team has. I watched for a while, as the Group C Nissan went around and around. There were only two guys manning the computer banks, so I walked up and asked what they were logging. The answer was that they were logging over 200 channels of data. I tried to think of 200 channels I would want to have logged. I got to about 30 and had to ask “what are you guys logging?” The Japanese engineer said “mostly engine parameters”. He said the engine block had twenty strain gauges cast into the block to measure stress in various sections of the block.
I thought that was very cool, but wanted to know what they were going to do with the data. The engineer said that, sometime in the future, a Nissan engineer was going to want to know what the stress was in an engine block. When that happened, the data would be there. He said that the cost of running a race car was so high, that collecting data while it was running reduced the cost of gaining that information later, when someone needed it. That impressed me with the long term thinking to collect data. Collect the data when you can, sort it out when you have time.
Any race team can do the same thing, on a smaller scale. Using a spread sheet, make a matrix with each column being a parameter in your setup. Include columns for driver comment, change, and result of the change. Each row will be a track outing. Start to log the result of each change you make to your car. In short order, you will have a history of each change made to your car.
Using the data sort function of the spread sheet, you can sort the data. For example, you can show every time you had exit understeer. Then, look at what you changed when you had that problem and what the result was. This will prevent you from making the same set up mistake twice. And, it will tell you what things have worked to correct an exit understeer in the past. It will keep you in the window of your knowledge. When testing, look at the data base and try some things that you have not tried before. This expands your understanding of the car and makes your data base more helpful at the track.
Using a data base program, like Access, takes it to the next level with tables, forms, and special queries to make the search results more meaningful and detailed. If you spend some time (it took me about 30 hours and I am NOT an Access expert), you can answer more questions. You can ask the database to show what we changed to reduce a mid corner oversteer in a third gear turn when we had tire pressures above 15 psi and a rear wing setting below 10 degrees.
Remember, the sooner you start collecting the data, the cheaper every mile you do becomes, because you are getting more value for your dollar. I can say that this has helped me a bunch over the years. My Daytona Prototype data base has over 900 changes logged in it. It just makes me a better guesser.
So, to recap…
1 – The driver is in charge of his setup. The engineer just makes the change he thinks will help him the most.
2 – Resist making changes until the driver really has a good feel for the car. It has to “talk to him” first.
3 – Physics does not change when you cross state lines. Go with what you know.
4 – The secondary effect of a change can be in the opposite of the primary effect. Never make a change without considering the secondary effect.
5 – Use a data base to increase your understanding of your car quicker and with more accuracy.
6 – If a change does not have the effect you thought it would, than you are missing some effect that you did not consider. There is not some weird phenomenon going on, you just don’t understand the circumstances of the particular situation.
7 – When in doubt, go back to your base setup and start over from there.
8 – Never copy another faster team’s setup. You need to know why yours does not work, so you can be better next time.
9 - If a change works the way you thought it would, you did not learn anything. You did become faster, which is always a good thing. But, when the change does not work as planned, you have a great chance to become smarter. Grab on to that and figure it out.
10 – When recording changes in your notes, write down why you made that change, your thinking on what you expect it to do, and why. Then, you can go back later and see what your thinking was for making that change and decide where your thinking was wrong. This may happen months later, as you get to know the car better, but it allows you to see where your mistake was, not just that it was a mistake.
OK, one more thing.
My friend, the driver/engineer mentor of mine, wrote these points down once about problem solving. Think about them. It may change your approach to finding the perfect set up.
1 – What’s right is right and everything else is wrong to some degree.
2 – What is REALLY happening here?
3 – Nothing happens for no reason.
4 – Everything is attributable.
5 – If X is true then Y must also be true. If I can’t prove that Y actually does as I predict, then I don’t know anything at all about X.
6 – What I am certain is correct can change instantly in the light of what is REALLY correct, whether I like it or not.
7 – Just because I don’t want to believe it, doesn’t make it wrong.
7a – Just because I want to believe it doesn’t make it right, either.
8 – Knowing what is wrong is every bit as important as knowing what is right.
9 – If it isn’t all the things you think it is, then it is something else. (Sherlock Holmes)
10 – You only know something if you can prove it. Everything else is “I suspect” or “I guess” or “I wonder if” or “it is my theory that...”
11 – The right answer is still the right answer even if you didn’t think of it.
12 – The right answer is still the right answer even if you don’t have any idea of why it works... but find out later for sure, because the underlying principles will always apply.
13 – Asking other people for answers is perfectly acceptable, as long as you never believe them.
Friday, February 5, 2010
Setup Sheets, Part 4
We'll wrap up the series on setup sheets with some specific thoughts about parts of the sheet's content. For the most part, you can refer to the basic sports prototype sheet from Part 1, reproduced below.
Sports Car Prototype Setup PDF
Tolerances
You and the crew need a mutual understanding of the tolerances for any given adjustment. There are two pieces to this. The first is your knowledge, as engineer, of what constitutes a significant change in any adjustment. The tolerance, of course, should be tight enough to avoid changing the car by missing the target. The second is the crew's ability to hit your target and your tolerances, either because of finite steps for some adjustments, or because of inability to either measure or adjust tighter than some tolerance. For example, if you call out a toe-out setting to a tolerance of .002", you are way tighter than adjustment repeatability, measurement accuracy, and large enough difference to effect the car.
Specific hardware
When you know it, call out the specific part number, hole number, shim thickness, or other part or adjustment specification that should lead to a specific setup target. The setup sheet above, for instance, is for a car with front camber slugs, rear camber shims, caster shims, and specific holes for wing adjustment. In each case, both the setup target and the hardware spec are called out on the sheet. In the event that the crew can't hit the setup target by following the hardware spec, something is either wrong or misunderstood, and discussion will be required.
Standard setup conditions
For the sake of repeatability and comparison, it helps to always set up with the certain setup conditions always the same. Tire pressures, driver weight, and fuel load are chief among these. If you're not using "setup wheels", the devoted metal fixtures that replace wheels and tires, then you should always use the same set of actual tires and wheels. Actual running conditions for the start of the session will vary. For instance, we'll set up at typical hot tire pressure, but start the session lower than that. Starting fuel load may be more or less than setup. You get the idea.
Actual vs. target
Some fields on the sheet may have spaces for actual vs. target. This is helpful if you want to record what you intended vs. how the car actually ran. The difference could be from tolerances, measurement precision, or simply error.
Corner weights
If you have a good grasp of the total weight and the percentages for front and left side, you should be able to predict the actual corner weights for any given cross weight within a couple of pounds. I tell my crew that, if they miss the target corner weights by more than a certain margin, they should contact me so we can figure out what's up.
Wings
Lots of folks call out wing angle as relative to the reference plane of the car, which is usually defined as some part of the underfloor. This seems simple enough. Time and again, though, it seems that zeroing the Smart Level against the floor isn't a totally repeatable process. I've come to call out wing angles as absolute, versus the ground. The results, for me, are more repeatable.
I learned this the hard way, in my early days in IRL. We had a flawed zero against the top of the tub floor inside the car. For part of the weekend, we struggled with lack of grip as we ran less wing angle than the series actually would have permitted us to use.
That about does it for setup sheets. Hope you've enjoyed it.
Sports Car Prototype Setup PDF
Tolerances
You and the crew need a mutual understanding of the tolerances for any given adjustment. There are two pieces to this. The first is your knowledge, as engineer, of what constitutes a significant change in any adjustment. The tolerance, of course, should be tight enough to avoid changing the car by missing the target. The second is the crew's ability to hit your target and your tolerances, either because of finite steps for some adjustments, or because of inability to either measure or adjust tighter than some tolerance. For example, if you call out a toe-out setting to a tolerance of .002", you are way tighter than adjustment repeatability, measurement accuracy, and large enough difference to effect the car.
Specific hardware
When you know it, call out the specific part number, hole number, shim thickness, or other part or adjustment specification that should lead to a specific setup target. The setup sheet above, for instance, is for a car with front camber slugs, rear camber shims, caster shims, and specific holes for wing adjustment. In each case, both the setup target and the hardware spec are called out on the sheet. In the event that the crew can't hit the setup target by following the hardware spec, something is either wrong or misunderstood, and discussion will be required.
Standard setup conditions
For the sake of repeatability and comparison, it helps to always set up with the certain setup conditions always the same. Tire pressures, driver weight, and fuel load are chief among these. If you're not using "setup wheels", the devoted metal fixtures that replace wheels and tires, then you should always use the same set of actual tires and wheels. Actual running conditions for the start of the session will vary. For instance, we'll set up at typical hot tire pressure, but start the session lower than that. Starting fuel load may be more or less than setup. You get the idea.
Actual vs. target
Some fields on the sheet may have spaces for actual vs. target. This is helpful if you want to record what you intended vs. how the car actually ran. The difference could be from tolerances, measurement precision, or simply error.
Corner weights
If you have a good grasp of the total weight and the percentages for front and left side, you should be able to predict the actual corner weights for any given cross weight within a couple of pounds. I tell my crew that, if they miss the target corner weights by more than a certain margin, they should contact me so we can figure out what's up.
Wings
Lots of folks call out wing angle as relative to the reference plane of the car, which is usually defined as some part of the underfloor. This seems simple enough. Time and again, though, it seems that zeroing the Smart Level against the floor isn't a totally repeatable process. I've come to call out wing angles as absolute, versus the ground. The results, for me, are more repeatable.
I learned this the hard way, in my early days in IRL. We had a flawed zero against the top of the tub floor inside the car. For part of the weekend, we struggled with lack of grip as we ran less wing angle than the series actually would have permitted us to use.
That about does it for setup sheets. Hope you've enjoyed it.
Monday, January 4, 2010
Setup sheets, Part 3
We now veer from the content of setup sheets to focus instead on using them. Since the setup sheet is primarily a communication tool, that will be the theme of this post. The final post will cover specific content items on the sheet.
What happens on the first usage of a new setup sheet?
Some teams want you, as a newly-arrived engineer, to use their existing setup sheet. OK, then get on with it. If you need to make changes to the sheet, phase them in as needed, with agreement of all concerned. No surprises allowed.
Other teams will allow you to use your own format. Or, maybe, you've decided to rework the existing format. In either case, it's absolutely essential to sit down with the crew and review the format and content of the setup sheet. The compressed layout needed to convey so much information will almost certainly lead to misunderstandings if you bypass this step. Remember, communicate!
When does the setup sheet need to be done?
Let's talk about the pre-event sheet first. You could wait until a day or two before the car goes on the setup pad. You'll probably only get one or two opportunities to do that, though, before someone gives strong verbal feedback, or worse.
On the engineering side, we often want to review notes and data from the previous event before deciding how to set up for the next event. Maybe we've got wind tunnel or 7-post testing coming soon, too. Maybe some significant analysis or simulation work will reach conclusions soon. All this work could have a significant effect on the setup, and not just on adjustments. We may need to make (and test-fit) aero or chassis parts, revalve shocks, buy parts, and more.
Meanwhile, the team typically has a limited amount of time to turn the car around. There will be a finite amount of manpower available to deal with changes. And, all this will usually happen on a schedule without much wiggle room.
What happens on the first usage of a new setup sheet?
Some teams want you, as a newly-arrived engineer, to use their existing setup sheet. OK, then get on with it. If you need to make changes to the sheet, phase them in as needed, with agreement of all concerned. No surprises allowed.
Other teams will allow you to use your own format. Or, maybe, you've decided to rework the existing format. In either case, it's absolutely essential to sit down with the crew and review the format and content of the setup sheet. The compressed layout needed to convey so much information will almost certainly lead to misunderstandings if you bypass this step. Remember, communicate!
When does the setup sheet need to be done?
Let's talk about the pre-event sheet first. You could wait until a day or two before the car goes on the setup pad. You'll probably only get one or two opportunities to do that, though, before someone gives strong verbal feedback, or worse.
On the engineering side, we often want to review notes and data from the previous event before deciding how to set up for the next event. Maybe we've got wind tunnel or 7-post testing coming soon, too. Maybe some significant analysis or simulation work will reach conclusions soon. All this work could have a significant effect on the setup, and not just on adjustments. We may need to make (and test-fit) aero or chassis parts, revalve shocks, buy parts, and more.
Meanwhile, the team typically has a limited amount of time to turn the car around. There will be a finite amount of manpower available to deal with changes. And, all this will usually happen on a schedule without much wiggle room.
It isn't really possible to say exactly how long ahead of the next event you need to be done. The correct answer is "it depends".
- Are you taking the same car to the next race as the previous one?
- How time-consuming are the changes envisioned, beyond simple parts swapping and adjustments?
- When does the car need to go in the trailer? How far ahead of that must the car be set up?
- How important are the non-adjustment changes, in terms of performance potential? How sure are you of their value, and how sure is the crew of their ability to complete them?
- How "non-routine" will the setup pad work be?
I can't, repeat cannot, overemphasize the importance of communication with the crew. If you envision changes, then you absolutely must discuss them well in advance of actually asking for them. Jointly review their probability of actually happening, the work required to accomplish them, when you can realistically have them defined, and what the possible pitfalls might be. Then, jointly decide on a plan.
A few thoughts:
- The optimal time for the setup sheet to arrive is just before reassembly of the car begins, and with enough lead time for all the changes.
- It may help to produce a preliminary setup, to be revised as engineering work finishes. Be advised, this approach is full of pitfalls and requires a clear understanding from everyone involved.
- When the races are coming hot and heavy, maybe on back-to-back weekends, you're likely to have to prepare setups for several races ahead of yourself.
Here's a clue - when the crew is consistently asking you for a setup, then either you are too late producing them, or they want them too far in advance. Talk with each other, figure out the situation, and fix it.
How to deliver the setup sheet?
I prefer to add a "cover memo" to draw attention to anything important. Maybe you're changing something that has remained constant for a long time. Maybe there are exercises to be done on the setup pad, or some specific instructions on how to do part of the setup. It's easy to have this discussion verbally, but there's always some risk of misunderstanding.
Now, what about between sessions at the track?
The basic situation between sessions is much like the pre-event setup, only compressed in time. As an engineer, you want to debrief the driver, look at some data from the car, punch some numbers into the calculator or software, and maybe crank out a sim run or two. The crew wants to get the car ready for the next session. Here's how you do it:
On the way back to the paddock or the garage, tell the crew chief whether you need a setdown. If not, then he can start on the maintenance issues while you work on the setup. Tell him what changes you're considering, if it entails a significant job (for instance, changing brake master cylinders can be a challenge on some cars), and agree on whether the change needs to happen now or can wait until the end of the day. Changing ratios in the gearbox, for instance, typically needs some lead time - you know you're going to do it, even if you don't know what ratios to install yet. Finally, agree on when he needs your final answer on the changes you just discussed.
Spend the rest of the walk back deciding if you can call out the setup changes NOW. If you can, you are now officially a hero.
Put out a job list for the crew, DAG, shock guy, etc. immediately after arriving at the paddock or garage. No ifs, ands, or buts.
Get into your job deciding on the setup changes with energy and focus. Debrief, analyze, and decide. Now. Take it as a personal challenge to avoid agonizing over decisions. This is going to sound brutal, but if you repeatly have to agonize over the setup, then you either aren't ready for the job or you aren't decisive. Either that, or there's something you haven't learned about the car and desperately need to figure out.
If you spend two hours fussing over the setup, with the crew waiting on you and the next session looming, only to decide on a big change like tearing into the gearbox (stressing the crew) or a trivial one like a minor spring change (insulting the crew), you are not a hero, and will not be able to repeat this behavior forever - unless you and the driver win a lot of races!
Once you've decided, communicate your setup changes clearly and concisely. I found out, the hard way, that a new full setup sheet IS NOT the way to do this between sessions. There is too much information on it for the crew to pick through, while searching for the changes. I simply provide a list of the changes. Sometimes the setup worksheet from the previous post in this series is a good way to do this, sometimes it goes on the job list. It's best to discuss it with the crew chief, no matter what format you use. Avoiding misunderstandings is the goal.
That's about it for today.
Subscribe to:
Posts (Atom)
