I was thinking about this the other day when I was arguing with somebody about something like this. So often people think that a chassis parameter is very important because they had experience with that parameter being way off. That alone is not enough to earn a parameter it's standing. By that standard tire pressure might be the most important because you can't go fast on a flat, or shocks might be the most important because without damping the car would be undrivable, or alignment because if it's grossly off the car would be undriveable.

Your baseline for evaluation cannot include factors that are way off - those should be considered errata and be corrected.

So what is important? The tires need a nice temp profile across the width of the tread, and the tires need to see minimized vertical tire force variation. This last point isn't meant to suggest that we want to minimize weight transfer - weight transfer is dependent on longitudinal and lateral acceleration and the basic design of the vehicle (track, wheelbase, center of gravity height, and sprung weight). We want to minimize disruptive force variation - an example of which would be caused by excessive rebound damping reducing vertical load on a tire under lateral force when we don't want it, or excessive compression damping increasing vertical load (and necessarily decreasing vertical load on the other diagonal). We have to weigh such effects against other parts of the lap where that amount of rebound or compression damping might be helpful. Now consider a matrix with dozens of factors and even more specific points on a lap for evaluation and weighting.

There is no escaping the finality of the process imposed by the stopwatch. Much can surely be done without a 7-Post rig, if you think really hard about things.

Scott, who thinks that photo of Ron Tauranac in the latest RaceCar Engineering is a frameable classic.....


[Modified by RR98ITR, 11:38 AM 1/29/2002]

 

I think that you are asking a question that doesn't have an answer, other than "it depends".

The first step toward understanding this situation might be to lump your variables into "types", to make the issues more manageable. (But only a little more so.) Nonscientific terms follow...

"Input" Variables - settings that you can control (e.g. toe, camber, tire pressure, etc.) There may be subsets of these - variables that are pretty much constant, and variables that shouldn't change but do (whack a curb).

"Environmental" Variables - things that you can't control (e.g. ambient air temperature, track surface temp, etc.) Again, these may or may not change over the period of time in which you are interested.

For special consideration, "Interdependent" Variables - those which change as other variables change. These may be direct (track temp goes up, tire temp goes up), inverse (more camber -> lower temp on outside edge of tire) or some wacky combination thereof. In reality, pretty much all of the variables involved are interactive to SOME degree, which makes setting up your chassis such a tail-chaser.

"Output" Variables - those that you are trying to control but cannot change directly? Tire temperature is a good intermediate one but lateral and longitudinal grip at the contact patch is what it is all about. This is measured indirectly with the stop watch.

All of these influences then act over time, on a lap or over the duration of a race - making compromises necessary. More front camber might make for a quicker car mid-corner but hurt acceleration coming off of it, for example. This is why different set-ups work better on different tracks. Driving style has to be consired as well. Change tire brand or size and you start from scratch. Etc. Etc.

You haven't even considered aero issues yet - more rake in the car might make more aero downforce but compromize front roll center height. VW Rabbits w/stock suspension links run nose-high for this reason.

I can draw a matrix with two, three, and even four dimensions (that's a cool one) but how about 17? Yikes.

Kirk

 

Kirk,

I'm not actually asking a question. Rather, I'm discussing a root of major fallacy. Thanks for joining in. The underlying supposition that some solution can be reached thru a descriptive matrix is the heart of the sporadically ongoing series in RaceCar Engineering titled "Design of Experiment". There were a couple of letters in the current issue, and the author in responding was a little "cheeky". So we wait......and kill one weekend after another gutting our cars, cutting up our hands, cussing at our beloved G3's (I once told myself I'd never swear at such a fine car....then the damn thing bit me).

Scott, who technically only cursed the individual components involved in the pain and bloodshed......

 

I say the most important part of chassis setup is still the part that sits in the drivers seat. More adjustments can be made there than anywhere else.

An "ill handling car" in the hands of one driver may well be heaven in the hands of another.

So, if one setup can be good and bad, then can there ever be a perfect one?

Scott

 

Scott,

In the movie Gattaca, the character played by Gore Vidal says something like: "nobody exceeds their potential - if they do we made a mistake and we revise their potential."

Leaving aside literary inflation reserved for the likes of Gilles Villeneuve (which is not the same thing as discounting his heart), a driver can't pull more out of a car than it's capable of giving up. This also effectively means that whatever you've got, up to a point, a better driver can probably come along and get more out of it than you.

It is relatively rare for two similarly adaptable and skilled drivers to fail to reach the same performance with an ill handling car. Typically it will be the less adaptable and skilled driver who copes less well with an ill handling car. No driver would casually compete against a worthy peer in an inferior car and expect to overcome them thru skill alone. Funny, I had a friendly argument over "winning attitude" that touched on just this thing at my clubs banquet this past weekend. Take two drivers of equal skill and put them in cars of significantly differing capability. The driver of the inferior car will not be displaying a winning attitude if he thinks he's going to win - he will be displaying an unrealistic attitude. He can hope for his competitor to make errors or suffer problems, but that's it.

Within a fairly narrow range for any given configuration there is a small tolerance for the sort of variation that might fit in with the idea of driver preference. Any large difference is likely to lead to diminished total performance.

Perfection is elusive yes. But that's not the same as saying a pig and a racehorse are just as capable in the right hands.

Scott, who thinks that the driver is very important of course.....but that's not what this is about...



[Modified by RR98ITR, 2:29 PM 1/29/2002]

 

Okay - in the spirit of this conversation...

...there is, setting all of the human factors aside for a moment, some physical (theoretical) maximum grip at the four contact patches. The sum of these is determined by the vertical load on each, the coefficient of friction of the surface, their area, and molecular and mechanical grip of the tire materails. If the goal was to, on some kind of test rig, maximize this grip in steady-state situations, the job would be tons easier than it is. Attempting to maximize grip in this fashion is a step in the right direction - for a real race car on a real track with a real driver - but only a small one.

Having thought about it for a bit, the real magic (both mechanical and from a driver's standpoint) is maximizing the efficiency of the transitions between accelerative states - longitudinal (braking/acceleration) and lateral (cornering). That's where driver talent comes in. An inherently unstable car will transition very quickly from its steady-state, zero-rotational-inertia trip down the straight, to something like a steady-state cornering situation but it takes a talented shoe to live with this. J.P. Montoya is reputed to have some of the fastest hands in the world (there was an in-car shot of him catching his CART ride with full-lock steering on a superspeedway), which gives him the ability to work with a "fast" car.

Kirk

 

We can easily see the challenge posed by a high limit with a steep approach and a pin point top. Our Honda tin tops will generally have a less steep approach - up to those 1200F/1600R rate limits I've heard about. I've often thought the challenge of driving something really sloppy with terribly slow transients was almost harder to deal with.

My mentioning of off-track testing apparatus is not meant to suggest that I think it's the be all and end all. But any number of important engineers have commented on the importance of such rigs. My point was that we can achieve much by being sensitive drivers and relying on stopwatches - which as it happens is what most of us can afford.

And the transitions between the accelerative states is precisely where the attention to dampers is focused.

My favorite part of my motorsport is excercising my abilities behind the wheel. I try to take advantage of the time off the track to maximize the potential results on track. Both efforts come together on track - that's the proof and or payoff.

Scott, who thinks genius drivers prefer well set up cars so their talents are not wasted....

 

Where you say steady state you might not mean that. 7-Post rigs are about anything but steady state. Nearly every aspect of track surface and driver inputs can be mapped to the rams.

Scott, who while driving home tonite briefly became one with all five of his cars axies simultaneously......oh, joy.....

 

I wasn't explicitly thinking of "table" testing as it had been mentioned earlier. I should have worked on that transition myself (the one in my writing, that is). Replace "test rig" with "skidpad" and you get the same kind of idea, maybe? I did mean steady state - which could be simulated on a table but, of course, a whole mess of loads could as well.

Good point about shocks, BTW - that is where most of the voodoo is.

K

 

To set up a car for "idea" handling and "maximum" performance as stated here means there has to be a "perfect" place to do it. Maybe you can do it in a lab, but what good is that (other than for conversation).

No point in having a car if it isn't driven (I find auto museums to be sad places).

Now once you go out on track, how many variables do you have? The weather can change bewteen laps, even during laps, some pavement may be perfectly smooth other spots may be rough or even be a different kind (concrete patches for example).

If you set the car up to tear around right hand turns but then hit an off camber left, where are you? In the grass?

So, with all of those things taken into consideration, what is the best setup? And, will that same setup be right for two drivers?

Scott


[Modified by celica73, 1:33 PM 1/30/2002]

 

Wow, maybe you're right. There's no point in trying to set up a car - better to just randomly select spring rates, damping curves, alignment settings, tire pressures, etc, and let the driver work his magic. And if the lap times don't reflect what other drivers do with the same car it's clearly the driver that's failing.

Scott, who may have to look into Museums as places for conversation about being off in the grass....makes me wonder what in the hell you think I'm talking about here because it's not familiar to me at all.....

 

Ahh, I see where you are coming from now, I fully understand your logic (or lack there-of).

Yes, there is a most important part of a setup. Obviously it is the camber angle.

Maximum grip on a NON MOVING car (perfect setting) is ZERO. This maximizes the chance that the car won't slide off the level showroom floor.

Since you want to deal with statics in a dynamic environment, we will do it scientifically. We will set all veriables to zero except for one - gravity.

And, the best way to maximize the static friction between a car and a surface is to maximize the contact patch, so set the camber to zero and get back to your day job.

Remember, that is the scientific method. 1 equation 1 unknown, only change one variable at a time.

Maybe we can commision a panel of engineers to examine our non moving car and tell us how we can make it not move any better.

Scott (the onther one in this argument, who is just as big of a pain in the *** as the one that started it)

 

OK, have given this some extra thought in the more rational sense, I think I have decided there is a "most important part of chassis setup."

Since the question (as phrased) is really looking for a single thing, I will say it is the tires.

Lets go back to my scientific method (the most useful part of the last ill concieved post). If we hold all other things costant, the tire choice will likely make the biggest difference.

In autocross, what tire do you choose? Probably Hoosiers. Why not just do an alignment and call it good? Because the Hoosiers make the biggest difference.

This is more subtle with an F1 car where now you have a fully prepped car with race tires to begin with. But, I still maintin proper tire selection (or management) will be the fastest way around the course.

Use the wrong compound and blister the tires, then you are in trouble. Put wets on and it dries out and you are in trouble. More than once we have all seen someone on wet tires drive the WRONG line just to find water on a drying track. So, tire selection is critical.

Add pit stops to a race and it is even more critical. How fast are you going while you are getting tires changed? That's right, not very fast.

Lets go out and buy a new RSX, you change any ONE setting - camber for example and run your best lap. I get the same car but change the tires to Hoosiers.

I win.

Scott

 

Scott,

You really need to work on your reading comprehension - your responses are not reflective of comprehension of what I actually wrote - and on more than one level.

My initial post was about how some people think that there is one most important factor, and I took that topic to the elements at the contact patch. I specifically described what kind of data to throw out and why - and you in turn made your last argument using precisely such fallacious reasoning.

I am also amused that you've chosen to respond in what appears to be a complete ignorance of what a 7-Post rig is and what it does, and suggest that I'm mired in some form of static analysis of a dynamic phenomenon.

Nothing in what I wrote suggested that I view chassis tuning as anything other than an iterative search for the optimal compomise during one brief moment in time and place in space - and a moving one at that.

Good for you on the big win in your fixed race - you couldn't have done a better job of making my argument for me thru your near perfect execution of miscomprehension and psueduo scientific jackassery. And whoopie - you've had some exposure to Linear Algebra.

Scott, who knods in acknowledgement of a bigger pain in the ***.....and who wonders why this other Scott exhibits such distaste for posts of a technical nature as reflected in his own words which this Scott read carefully and accurately....maybe you should focus on your day job and not read or respond to my posts anymore....and don't read RaceCar Engineering or RaceTech either - you might learn something.

 

In my own previous post there is an unintentional opportunity for a reader to go wrong:


I meant to say "And if the lap times don't reflect what other drivers do with the same car (the type or model - but theirs with their setup) it's clearly the driver that's failing."

I wrote that out of a perception that I am reading arguments that boil down to: chassis tuning is pointless because there is no optimal setup. I can't believe that any thinking driver who actually uses their car on a road course and studies motorsports science would make that argument, but maybe this Celica73 guys isn't such a person, but instead a troll of some peculiarly annoying kind.

Scott, who wishes the board weren't like this sometimes....

 

Scott(s) - this has suddenly gotten a lot less fun than it was when I first joined in. I don't want to play anymore.

Kirk

 

haha, you guys get pretty onery during the off-season. why don't we take a time-out and have dreams of racing on our sleeping mats, vrooom! i'll take 2 cookies with my milk please!

 

Scott-

You've posted the question (even though you say it wasn't a question), and others (me included) have tried to answer it, sometimes seriously and sometimes not.

For every answer you seem to have some sort of problem with what is written, yet in this whole proces I've not seen *your* thoughts on the subject.

Since you appear to be so smart, why not enlighten us? Instead of presenting us with the riddle, why not try *your* hand at a solution.

We know that a 7 post setup can be used to simulate the course, we all agree that in similar cars the better driver will win. In your first post you even went as far to say the car is already setup properly and we are now fine tuning the variables for the lowest elapsed time.

So, what are the variables that are the biggest? Does it matter? Take the car out of the lab and drive it. On a *real* track, that is what the lab is trying to simulate anyhow (even thoughit can't come close - not on the level you want).

In the end, I'll still pick my post about the tires being the most important variable (which is ironic becase that is the one you chose to ignore, or simply agree with it).

Of course you could then argue that changing the air pressure in the tire is actually changing the spring rate so I no longer have one variable (any anyone that is worth anything would make that argument).

So, Scott, what do you think is the most important variable?

Scott

Major edit at 8:33AM EST


[Modified by celica73, 1:33 PM 1/31/2002]

 

(Aside- what an amusing pissing contest we have here)

I don't have my article on hand, but I'll paraphrase from Sportscar magazine a few months ago. I believe the words of Mark Daddio explained something to the effect of "What are the chances that you will arrive at the track with a perfectly set up car? Zero. It is not the ability to derive 'the perfect setup' but the ability to understand the goal of each change and apply them appropriately that matters."

To inject a little simple logic into this melee, take a look at your high school grades. If all your grades other than math and science courses were fertilizer, MIT is not going to take you into their engineering department regardless of how well you did in Physics. You need to be well-rounded. Barring similar extrema from the test matrix like "blown tires", it needs to be examined as to what the "group" of most important factors are. Many of the factors tie together, since as has been pointed out before, tire pressure changes not only affect the tread usage of the tire but also the effective spring rate. Perhaps it would be more effective to explore the interrelationships of test variables, rather than trying to find 1 single most important one that becomes irrelevant when other variables are not up to par as they should be.

That said, those 4 tires are the only contact between the vehicle and the road. Every other change you make to the car effects how weight is loaded, unloaded and distributed on the tires.

 

Again you provide proof that you can't read.

It is a question that others ask me from time to time and it is on the surface naive. That's why I wrote the title in the form of a quotation and used the word "like" as it would be used in a spoken sentence by certain familiar types.

And IT IS A RIDDLE.

With similar drivers the better car will win - that's why some of us try to make our cars work better (that and because better cars are more satisfying to drive).

I'm not interested in tires in the sense that you brought them up because that is not at all what I'm writing about. That question would read something like: "what is the one thing I can do to my stock xxxxxx to make it go around the autocross track faster?" That your suggestion of tires would likely be the best answer to that question is irrelevant here.

Try to think in terms of a spec class where only dampers, springs, bars, and alignment are free. Those variables are what are generally being referred to when the topic is "chassis setup". Things like tires and aero would be fixed. I wrote about extrema and fallacy - do you get it now?

Chassis tuning is an iterative process, and on each iteration one factor WILL have the highest sensitivity and response (some of us know this intuitively and are interested in the "Design of Experiment" series that I mentioned because it becomes harder to find improvement as you improve the car (because the variables are all interelated, and noise becomes hard to distinguish from data, and inconsistencies in tires, conditions and driver). In the end it will come down to the dampers. That's what you chase minimized Tire Force Variation with. That's where you find those little bits of grip that make you just a little faster than the other guys. I actually said that in my original post if you had known how to read it.

The number of iterations during a test session whether in the lab or on the track will be limited by time. Each iteration is judged by the laptime, the key test data acquired, and the drivers subjective feedback. Time and money impose a limit on everyone involved in such a process.


Again - what do you have against people trying to make their cars work better? I do take my car out and drive it, and I keep making it work better, and I keep improving as a driver, and I keep going faster. And I don't have a lab - I only have a brain, and I try to use it, and I share ideas that might be interesting on this board, and I enjoy some of the discussions I have (but not ones like this where people educated beyond their intelligence among others want to have a pissing contest).

Scott, who always enjoyed the Masters of "Kung Fu", and found the riddles interesting....that's all it is...