What is Handling?

[Ceilidh]

Hello Everyone,

As some of you may know, I'm just a guest here, as I own a GTI, not a TDI, and have only recently registered with your very fine forum. But I'm a friend of Peter (Pyce), and having received an email from Peter saying that he's going to consolidate some of our postings from the Vortex into something a bit more accessible to the folks here (the Vortex postings are scattered amongst a host of rambling threads), I thought I'd get things started with a little introduction to things pertaining to vehicle dynamics and suspensions.

Now, I don't know exactly what Peter has in mind, so the format of things might change as we go along, but perhaps a good way to start would be to clarify some oft-confusing terms so that everyone's on the same page. Hence I've called this thread "What is Handling?", and we'll use it to nail down what different handling goals can be, and what basic steps one can use to accomplish those goals. Content-wise, we'll wait for Peter's lead (and it'll probably take a few weeks to fully get things going), but for this first post, we'll take a quick and cursory look at an oft-discussed handling subcategory:

What is Understeer?

So let's get started!

* * * * * *

WHAT IS UNDERSTEER?

Back when Peter and I were posting on the Vortex, sooner or later we'd log on one day to a thread we'd been laboring on, and find something like the following:

Post 1: The understeer on my car is awful. How do I get rid of it?
Post 2: Do "X".
Post 3: Do "Y".
Post 4: I did X, and it didn't help at all.
Post 5: Well, I did X, and my understeer disappeared.
Post 6: You're all crazy. Our cars understeer, and you can't do anything about it.
Post 3a: Do "Y".
Post 7: I did Z, and it really helped reduce my understeer. But the car wants to spin now -- my friend's Subaru/Honda/BMW/etc. understeers even less than my car does, and it doesn't want to spin; what should I do to make my car like that too?
Post 8: I'm a racer, and I've done X, Z, A, B, C, D (etc., etc., etc.) to my car, and it handles like a dream!! It wouldn't be safe for any of you to do that to your cars, because you're not good enough drivers. But because I'm so skilled, I can take it. Let me tell you about an exciting incident I had last week, where my car would have crashed if it hadn't been for my lightning fast reflexes......
Post 3b: Why doesn't anyone do "Y"? I'm telling you, "Y" really works!!
Post 9: Hey Racer, that's so cool! My car has X, A, B, E, F, and a homebuilt G that I made out of two hoseclamps, an ironing board, and something my dog brought home last year. If I did D as well, do you think maybe that....
etc.
etc.
etc.



Anyway, besides convincing people like Peter and yours truly that our time is perhaps better spent doing virtually anything besides posting anymore on the Vortex, exchanges like the above usually resulted from disagreements on what "understeer" really means -- which is why we should clear it up from the start. To wit:

The problem with "Understeer" is that it has a very specific, narrow meaning in the vehicle dynamics handbooks, and a very broad definition in the popular press (especially on web forums and in most all aftermarket advertisements). Thus people complaining about understeer are often (through no fault of their own) really complaining about something that's only indirectly related to true understeer, whilst others offer solutions to "understeer" that are solutions to entirely different problems.

So what is "Understeer"?

Judging from web posts and advertisements, it looks like "understeer" for many people is a vehicle flaw that prevents the car from turning incisively into a corner; that numbs the steering and causes a rubbery feel just when one would most like to keenly sense the road-tire interaction; that causes the car to drift sideways in a corner, tires squealing and howling ineffectually; that reduces the car's agility and ability to dart & weave back and forth; and that causes a car to plunge straight ahead when a turn is taken too fast.

But from a vehicle dynamics perspective, however, understeer is directly related to only one of the above "flaws", and has only an indirect (and often weak) link to all of the others. Thus if you're trying to correct the above problems, and blithely follow tried & true solutions that are known to reduce understeer, you might notice very little "improvement" in all but one of the vehicle "flaws", and you might introduce a new set of problems you never expected.

To explain: from a vehicle dynamics perspective, "understeer" is a stability term -- if a car is subjected to a sideforce (caused by a gust of wind, a sloped roadbed, a pothole, cornering g's, magnetic attraction, an attack by crazed gerbils, whatever), it can do one of three things:

1) Despite being shoved sideways, it can proceed merrily on its way, in the same direction it was originally travelling;

2) It can turn away from the crazed gerbils, etc.;

3) Or it can turn towards the gerbils, windgust, etc.

#1 is the condition of Neutral steer; #2 is Understeer; and #3 is Oversteer.

The three steer conditions have stability implications because the moment a car begins to turn, it experiences an additional sideload due to cornering g's. In the Neutral case, the car doesn't turn, so there's no additional sideload. In the Understeer case, the cornering sideload "fights" the original sideload, and thus the amount of turning decays with time (e.g., if a gust of wind shoves the car to the left, an Understeering car begins to turn to the left; but then the (right-directed) centrifugal force from this leftward turn counteracts the original shove, and eventually the car straightens out). In the Oversteering case, the cornering sideload augments the original sideload, and the turn becomes tighter and tighter (e.g., when the wind shoves the car to the left, the car turns towards the right, which causes centrifugal force to shove still harder towards the left, which causes the rightward turn to tighten still further, etc....), until the car finally spins.

Think about the implicatons of the above for a moment!: if a car is set up to be truly Oversteering, then any sideforce -- any gust of wind, pothole, slope, crazed gerbil, etc. -- will cause the car to spiral into a spin, unless of course the driver intervenes in a timely fashion by turning the steering wheel in the right direction. Conversely, an Understeering car is inherently stable: when you shove it to one side, it turns away from the shove and then tries to straighten out, without ever wanting to spin. This Understeering situation is inherently much, much safer, and thus it should come as no surprise that essentially all passenger cars are designed to understeer: if they were not, then high-speed driving would be a continuous white-knuckle struggle to avoid spinning off the road -- even if the road is dead-arrow straight.

So to reiterate: All Passenger Cars Are Designed to Understeer, and that is A Very Very Good Thing.

Given, then, that Understeer is a Good Thing, why do people hate it? Here's where we get to the crux of the confusion: there are legitimate reasons to dislike understeer, and there are confusing reasons. Let's look at the legitimate reasons first:

Legitimate Reason #1:
Even many (and perhaps most) racecars are set up to understeer mildly -- but HEAVY understeer will cause problems when a car is driven at the limit. Understeer causes a car to turn away from the center of a corner, and thus a heavily-understeering race car will try to plow straight into the concrete walls/ tire walls/ gravel traps lining the outside corner of a racetrack, giving the driver a brilliant view of what he's about to hit. Racing drivers really dislike having such a view, and they would much prefer the car to be less stable, so that they can try to "pitch" it into the corner and adjust their trajectory. For this reason, a common racing quote is "Oversteer is where the passenger is scared; Understeer is where the driver is scared.", and a lot of race car tuning is spent to eliminate heavy understeer, especially on low-speed, tight corners, where such understeer tends to be most severe.

Legitimate Reason #2:
Similarly, excessively heavy understeer on the street can have us plowing off the road during rainstorms or snowstorms, with the wheel cranked hard over while the car charges straight ahead, and that's unpleasant too.

Legitimate Reason #3:
And then there are people like Peter, who like to be able to bring the rear end of the car around on occasion, just for fun.

There are other good reasons for disliking excessive understeer, and we can extend the above list quite a bit -- but notice something: nowhere in any of the discussion thus far have we said anything about steering response, or corner entry turn-in, or left-right darting agility, or road feel. Understeer IS indirectly related to these other characteristics, in that a heavily understeering car is a very stable car, and stability works against lightning reflexes, but the relation isn't direct. Indeed, it's possible to have an understeering car that has great reflexes, or conversely a fairly neutral car with comparatively sluggish reflexes (at various times in his experiments, Peter's car has probably sat in this latter category: a TDI with minimal understeer but relatively slow steering response). Agility and Understeer are in many respects two entirely different things.

Why does the above matter? Well, it matters because the chassis and suspension modifications you make to reduce understeer are not the same changes you make to directly improve agility and steering response, and in a fairly common extreme case, it's possible to so reduce understeer in a misguided (and failed) attempt at improving agility, that the car becomes dangerous to drive on the street (especially in poor weather conditions). This last predicament is one that several Vortexers have encountered, particularly when they installed ever-increasingly stiff rear antiroll bars on otherwise stock cars, in a futile attempt at increasing turn-in steering response via "reduced understeer".

As for what chassis mods work towards reducing understeer versus what mods work towards agility -- well, that'll be in future installments, depending upon what direction Peter wants to take things. Cheers til then!

- Ceilidh

 

[nicklockard]

Thank you. Thank you. Thank you!

Thank you for starting this discussion!

Conversely, an Understeering car is inherently stable: when you shove it to one side, it turns away from the shove and then tries to straighten out, without ever wanting to spin. This Understeering situation is inherently much, much safer, and thus it should come as no surprise that essentially all passenger cars are designed to understeer: if they were not, then high-speed driving would be a continuous white-knuckle struggle to avoid spinning off the road -- even if the road is dead-arrow straight.


(snipped)

Why does the above matter? Well, it matters because the chassis and suspension modifications you make to reduce understeer are not the same changes you make to directly improve agility and steering response, and in a fairly common extreme case, it's possible to so reduce understeer in a misguided (and failed) attempt at improving agility, that the car becomes dangerous to drive on the street (especially in poor weather conditions). This last predicament is one that several Vortexers have encountered, particularly when they installed ever-increasingly stiff rear antiroll bars on otherwise stock cars, in a futile attempt at increasing turn-in steering response via "reduced understeer".

(emphasis added)

I look forward to seeing this discussion develop and evolve. Thanks!

 

[BoosTDIt]

Yep... I too can't wait to learn and absorb even more information on the subjet than i already have through scattered posts here and Vortex!

And i can assure you the audience here is somewhat more mature (intelectualy not age) thus the topic will remain in its intended educational form.

And it'll be preserved in a better way here at TDIClub, rather than the gazzilion multi-branch thread giant Vortex...

 

[TdiAvenger]

page 1

 

[somepunk22]

Increasing agility.... Are you guys going to invert the dampers? Or maybe just the rears? In a earlier fsd post of Peter's, he quickly mentions this.

 

[peter pyce]

Hi Winston!

.....I'm just a guest here.....

We are all guests here, except the guy(s) who pays to keep this site going J

I am glad you decided to come and dedicate time and effort to put few things together, so they do not get lost in space. It had been incredible, what, three years now (?) since we got in touch, and so much had been written and said, it will be shame if it gets lost.

I particularly think your amazing write up on the “Stock” setup from that long thread is something worth reading for everyone who wants to learn little bit more about what a Stock VW is all about. I have been searching through the Internet for few years now, in attempt to learn as much as possible on suspension, and I have never seen any text of this magnitude that talks about a stock setup. Everyone is so, so busy with “reporting” handling characteristics on modified cars, but almost no one spends time to understand what the stock car is to begin with, especially on a case like the VW, which definitely is not a sport car and unless some big modifications are done – it will never be. Also, I particularly like the text that “nicklockard” quote above – it is the essential of where the trouble starts when modifying a car! The goal is not so clear and the results are just perceived feelings, meanwhile the new setup is a lot worse in so many other aspects, but that is either overlooked or simply not even detected due to pure lack of experience and knowledge.

So, let me go find the long posts on the Stock setup and perhaps we can simply copy-paste them and go over one more time and then perhaps use those as a base for some sound conversation. It had been more than two years ago, so I am sure we can add something to that master piece.

 

[peter pyce]

Let's continue with this post of Ceilidh's..... it is, funny enough, from exactly two years ago This is about scene-setting here with a little discussion of endgoals and driver types:

The following is a copy-paste, but will not use quotes, so it will be easier to refer/quote later. His post all the way to the end:

If you look closely at the posts to the suspension forum, you'll see that suspension enthusiasts tend to fall into several more-or-less clearly defined groups. We'll ignore those who want lowering for looks, as well as the people who simultaneously want pinpoint handling precision, a luxury ride, fantastic grip, and failsafe emergency behaviour (advice to these souls: save up for a nicer car). The remainder fall roughly into 5 categories (I'm grossly oversimplifying here!): there are track enthusiasts, divided between autocross and road-course aficionados, and then there are the roadies, whose 3 categories I'll explain in a moment. (As discussed in the earlier installment (#2), the track setups are considerably closer to "neutral" than are the road suspensions, and only the very skilled or commendably cautious should be running serious track/autocross cars on public roads.) We'll go through these categories one by one:

1) Track-Based Suspensions: Autocross

Autocrossers live in a flat, low-speed world with endless sharp corners. These characteristics -- the relative smoothness of the autocross tracks (e.g., parking lots with cones), the need for lightning-fast corner entry and left-right transitions between corners, and the low speeds necessitated by the constant tight turning -- define the autocross suspension. As we'll see in a later installment, there's probably more room for variety in a Golf/Jetta autocross suspension than anywhere else, and it's here that one might get away with the ultra stiff springs, big antiroll bars, and meaty wide tires that are the staple of common speed shops. That's not to say that the aforementioned parts are the fastest or best way to go, but at least in autocross they have a chance.

2) Track-Based: Road Course

Compared to an autocross course, a true road-racing track allows for much higher speeds and a much more flowing mode of travel. Agility in transitions starts to matter less than stability under trail-braking and an ability to put down power on corner exit. Although the surface of a good track is about as smooth as anything a car will drive on, the speeds are high enough to begin putting a premium on absorbing bumps while maintaining traction. The requirements of a good road-racing suspension start to converge on those of a good street suspension, and a dialed-back road-race setup (i.e., a road-race setup that's been given better ride comfort and more stabilizing understeer) can form the basis of a fast road car.

3) Street-Based: The Darter

"Darter" is not an official term, but it is descriptive. A darter is a person who wants his car to have the instantaneous, solid, no-roll response of a go-kart. "Handling" here equates to an ability to zig-zag through turns so suddenly that passengers lunge frantically for the grab handles and loose french fries go flying through the cabin, or to change lanes in a quick left-right snick-snick that supposedly looks cool but in fact is only dangerous -- it's sort of the Super Mario/ Nintendo view of the automotive world. I was a darter once, and a lot of sub-24 male car enthusiasts seem to start off in this category. Unfortunately, darters are rarely happy people (not least because, if darting is the goal, the Golf/Jetta IV is the wrong car). For the VW-loving darter, the car always rolls too much, the steering is always too slow, there's always too much understeer, and friends' Hondas and Integras are always better. Darters play an important role in the VW tuning ecosystem by single-handedly keeping the smaller tuning shops alive: it is the darters each year who go through enormous quantities of springs, shocks, bars, coilovers, urethane bushings, camber plates, rims, tires, spacers, spindles, strut tower braces, steering racks, spherical bearings, and general miscellaneous stuff, all guaranteed by the vendors to "tighten the handling while eliminating understeer!". There's nothing wrong with being a darter (as I said, I used to be one myself), but it's a frustrating existence if your car is a Golf/Jetta IV, and ultimately darters either gravitate towards autocross setups that don't work for the real world ("Selling complete suspension!! Wife having baby -- must sell!!"), or else they give up by selling their VW, or by becoming a Grand Tourer.... In any event, there is no suspension setup that is the ultimate Golf/Jetta IV street Darter -- if darting is the goal, you will need a different car.

4) Street-Based: The GT's

The acronym "GT" originally referred to Gran Turismo (Grand Touring) cars that were fast, comfortable, and effortless at rapidly covering ground on difficult, twisting roads (supposedly they developed when Europe still didn't have much in the way of superhighways).. And the GT version of the Golf/Jetta is what most non-darter street enthusiasts are looking for. If you fall in this group, you're looking for something that has a lot of grip, is reasonably comfortable, possesses good stability, and is relatively forgiving (so that you can converse with your significant other while zipping through the countryside). The setup for a good GT is essentially a dialed-back version of a good road-race car (with increased understeer and more comfort); the Shine SRSS is probably the exemplar here (although it apparently works well on the track as well), along with its softened permutations, and the GT enthusiast can essentially decide where on the ride vs. handling continuum he wants to be. We'll spend a good amount of time talking about the GT setup, when we eventually get to the appropriate installment.

5) Street-Based: Driving a Slow Car Fast

Finally, here is the smallest, least popular group, which I list primarily because I (Ceilidh) currently live here. A sad thing about aging is that some of us become steadily more boring and wimpy as the years go by, and eventually we resign ourselves to driving interminably behind the Volvo 240 with Delaware plates instead of searching for a way to pass it. When that happens, a "good" suspension becomes one that provides maximum feel, that responds somewhat to classical driving techniques even when driven at 3/10 to 5/10, and that possesses low enough limits (or more accurately, "perceived" limits) so that we get the occasional sensation we're actually driving. For people in this small category, a near stock setup with better damping (but skinny tires) is actually kind of nice.
So those are the categories. Each category has a different ultimate suspension solution (except for the Darters, who are doomed to disappointment), and a person wanting to modify his suspension had best figure out which category (or between which categories) is the one that really applies.

By the way, if you want to have some fun, go through this thread and try to guess which category the various vortexers currently live in. Yours truly (Ceilidh) started out as a Darter, went next to Road-Racing, before fear drove him to GT, and then boredom led him to Slow Car Fast. Peter (Pyce) apparently began as a GT, pushed that to an extreme, and is now slowly drifting in the direction of Slow Car Fast (though I think he's still in the GT realm).

Next installment: how the stock suspension works.

 

[IndigoBlueWagon]

Ah, slow car fast. I'd like to be the GT driver but reality puts me on crowded roads that makes me the third category.

And I learned about understeer and oversteer at an early age by learning to drive in various swing-axle Volkswagens and an Austin Healy Sprite. Talk about getting old. That was back in the days when not all cars understeered. Triumph Spitfire? Any Porsche?

I started out here as a darter, too. But now am on stock springs and Koni Reds on full soft. Life is good.

 

[oldpoopie]

Ahh. I'm #4. I've never had the opportunity to be a darter... I used to drive a '69 ford falcon, all stock. Then went to a '95 chevy tahoe.... Then to the golf with Shine. I've loved the shine setup from day one. The thing that amazes me is how much toe-in alignment can effect the twitchyness of the car. Point is, make sure your certain your toe alignment is set properly before judging any setup.

 

[JetPuf]

The thing that amazes me is how much toe-in alignment can effect the twitchyness of the car.

When I first put in my TC's I had a bit of toe in, Yeah it was twitchy... on the way to the alignment shop.

 

[Golf_GTDI]

Great stuff. I look forward to see where this goes. I'm happy to see that we may be able to look at many aspects of handling and how they interact. I for one think I know where I would like to be on this above list of drivers but I'm not sure where I actually am.

I hope that my own efforts may in some way add a little something to the common good and find that I am paying a bit of a price in that I don't have a system that I'm all that happy with. Its also interesting that we bring up the stock system which as I recall was very stable but rather slow with minimal feel.

I could talk about what I feel now but I think that at this point I have learned enough to know that I know very little. I think I would have to try dozens of dampers and spring rates and many many tire combos before I could get close to a decent guess as to whats going on exactly.

I hope that as a group we can get a little closer to an understanding of what is going on with our cars and how we can get the effects that we want with minimal headache and less than endless testing.

 

[BoosTDIt]

A bit out of topic (Mk4's) but...i think the Mk2's are GREAT Darters
I know i had a few...lol

I'm in the Slow car Fast...aiming for GT ... but only the right type of GT... that's why i'll remain in Fast for a while

 

[Ceilidh]

Understeer, continued

Greetings again(!)

I think Peter's going to cut & paste a little more in a short while, so let me wrap up the understeer discusssion:

When we last left it, we had shown how "understeer" is a stability criterion, and we discussed how a car lacking basic understeer cannot be driven at high speeds, even in a straight line on a level road. Given that, we now ask: if all cars understeer, how is it that so many of them (e.g., mid-1960s Porsches) are said to oversteer?

The answer to the above question is one that most if not everyone on this forum already knows, but we'll discuss it for completeness: whatever the inherent balance of a car (whether it be neutral, understeering, or oversteering), the balance can be shifted -- to a considerable extent -- by what the driver does. Or in plainer English: a mildly understeering car can easily be made to oversteer by a skilled (or completely clueless!) driver, and the same car can be made to imitate a hopeless, understeering pig.

(Slight digression: there's a book somewhere out there (alas, I've no idea of the title or author, as I just leafed through it one day in a Philadelphia bookstore...) that showed Formula One cars going through a corner in a 1950s Grand Prix race. In that montage of photos, you could see Mike Hawthorn (British, wore a tie while racing!) muscling his front-engine Ferrari through the corner on successive laps: on one lap he was understeering, on another there was spectacular oversteer, on the next he was heavily understeering again, then still later he was in a neutral 4-wheel drift -- same car, same driver, same corner, same tires, same setup, almost the same fuel load from shot to shot. It's not that the car was changing; instead, Hawthorn (a gifted but inconsistent driver) was having trouble with his Ferrari's handling that day, and was trying different techniques in an attempt at finding a good, fast groove...)

The techniques that cause a car to under- or oversteer are the things you learn in a good driver's school or even in a good racedriving book, and we won't go into them here; instead, we'll just mention that many of the techniques hinge upon inducing fore & aft weight transfer via acceleration and braking, with the "heavier" end of the car generally winding up with more traction. What's important for this discussion is that this weight transfer has an almost unbelievably powerful effect on the car's handling balance -- regardless of whether or not the weight transfer is deliberate -- and it is very difficult to drive a car without causing these weight transfers to occur continually -- for good or ill.

It is because of the above situation -- that cars are extraordinarily sensitive to weight transfer, and weight transfer is extremely easy to invoke in the wrong way -- that street vehicles are not set up like race cars....

[Digressionary note #2: many car magazines will rave about some mondo sports car having "race car handling!!!!" -- but if you follow and read these magazines very carefully, you'll find that most auto journalists have never driven a true race car at speed on a race track; the ones that do (e.g., Peter Egan at Road & Track is a Formula Ford race driver) never, ever claim that a street vehicle (barring something like an Enzo) feels or handles like a race car -- the comparison is just absurd!]

...With race cars, the small amount of built-in understeer can be easily overcome by small amounts of forward weight transfer, and it's the driver's job to feed in that forward transfer only at appropriate times; if the driver screws up (e.g., as by reducing engine power midway through a corner), the car generally spins like a top. In contrast, a passenger car has to be set up so that it rarely or never spins, and in particular it must not spin when people do things that are "normal" from a human psychology point of view, but which are suicidal from a racecar point of view. The list of these things is fairly long, but a classic example is what people do when they enter a corner too fast, and the car begins to swing wide: the "normal", human thing to do is to slow down (i.e., by easing off the throttle and perhaps even jumping onto the brakes), and this sort of action will shift a car strongly towards oversteer (translation: "Your honour, the plaintiff's husband drove around the corner, and the car just went out of control!! Clearly the defendant has placed a dangerously defective product on the marketplace, and in recompense for my client's mental anguish and lost future economic support, not to mention the suffering of her children who must now grow up without a father, the defendant must pay at minimum $2.4 million in damages..."). To keep a car stable under these situations, one normally has to build in enough intrinsic understeer to keep oversteer from developing at all; even there, a dedicated fool can still accidentally invoke oversteer for a brief time, but said fool will usually be subsequently saved by the car's stability when he completely abandons any attempt at controlling the car.

[Peter, can you find that Vortex post we put up this year, the one where we cut & pasted a Brake Forum post from a fellow who was saved by understeer, but who thought it was his driving skill?]


IMPLICATIONS

How does the above affect us with our VWs? Well there are a few salient points to keep in mind:

1) Our cars are not designed to be race cars, and they have been given a lot of understeer. This understeer exists so that ordinary people can drive these cars without much fear of spinning in normal traffic situations. If we take away this understeer, we reduce our safety margin. For the non-racers amongst us, this reduction is not a good idea -- particularly since (as we will eventually get to, sometime in the next few weeks) most people who think they want to reduce understeer would probably be happy with increasing steering response and agility, while leaving understeer unchanged.

If there are any under-24 males reading these lines, I can guess what many of you will be thinking! You'll be thinking "A safety margin is well and good for the soccer moms of the world, but I can get by with less, 'cause I'm a better-than-average driver, with faster reflexes, more experience (Hey, I've AUTOCROSSED!!), and I'll be focusing on my driving, instead of talking on a cell phone with my brain half dead, etc., etc., etc....". And for you folks who think this way, well, how can I phrase this? .... Let's just say that I was an under-24 male once, and I thought exactly that way....and I was completely wrong! If you haven't done a lot of driving at near-race speeds on a race track, in a race car, you might be better-than-average, you might have faster reflexes, and you might be better focused -- but in terms of your critical automatic-responses in a road emergency (i.e., what your hands and feet will do without your thinking about it), you're not much better off than the cellphone-talking soccer mom when Something Bad suddenly happens. No, I don't expect you to listen to me on this! (anymore than I listened to people saying the same thing -- it comes with being Young and Male), but maybe a girlfriend or Mom or somebody will read these lines over your shoulder; in any case, do please leave the understeer in place, unless you really have a reason to reduce it!

2) One thing the aftermarket shares with the OEM manufacturers is a strong desire to not be sued out of existence by the heirs and executors of deceased former customers. As a result, many if not most of the "performance handling" modifications out there do not reduce understeer very much at all; indeed, many of them actually increase it. Mods in this (enormous) category include (and we'll explain why & how in a later installment):

A) "Sport springs" that drop the car onto front bumpstops.
B) "Sport dampers" that degrade front roadholding while increasing darter-like agility
C) "Sport Low-Profile Tires" that are more camber-sensitive than stock

Items like the above will sometimes even slow a car's laptimes around a race track -- but they'll feel very "racy" (at least for certain people) while preserving or enhancing the stock understeer (which is good, as it keeps the paying customers alive).

(Note: there are mods that DO reduce understeer (one of them almost got Peter killed on a mountain road), and we'll discuss them eventually!)

3) And finally (at least for this installment, as it's getting late): since cars in general are so sensitive to weight transfer, and weight transfer is so easy to invoke (rightly or wrongly), an investment in a good driving course (or sometimes even in a good driving book -- a printed non-internet one!) is in many ways the most effective way of reducing or eliminating understeer. Case in point: during my own darter phase, I thought the ability to zig zag violently down a straight road, on a steady throttle, at fairly high speed, was the mark of an excellent car -- and it took me 2 years and much money & scraped knuckles to turn my poor little MGB into a car that could do that. It was only then that I learned that this much-modified car was now a frightful pig, and that I could get an equally incisive corner-entry turn-in with a completely stock MGB -- with much better overall handling characteristics -- simply by trail-braking a little at the beginning of the curve.

(And in passing, I guess that's why Peter and I are trying to write this thread: we both (independently) wasted a lot of time and money pursuing handling chimeras (and we both independently spun off into the weeds before we figured out what we were up to), and it'd be nice if we could save some of you some of our troubles!)

Ok, that's all for now -- Peter will likely cut & paste a bit, and then we'll resume next week or month. Cheers again!

- Ceilidh

 

[peter pyce]

..... Peter, can you find that Vortex post we put up this year, the one where we cut & pasted a Brake Forum post from a fellow who was saved by understeer, but who thought it was his driving skill? ......

Oh, that was one of the greatest! I have it bookmarked

Here it is, yours all the way to the end:

It's been raining for a long time here in CT, but this is the first time this has happened. Coming home today I was going around a corner and saw traffic ahead.. I lightly applied the brakes, and immediately started fishtailing.. Almost hit the car to my left, then came back around as I countersteered.. Almost hit a telephone pole to the right as my rear end surfed up along the curb.. I was able to let up on the brakes and straighten her out just in time.. Thank god for autoxing and learning what to do in such situations.. Otherwise I would've surely wrecked.
Now for the questions:
- I know our brakes are rear biased, but why would the front ABS engage without the rear?? I felt the pedal pulsating so I initially thought I would be able to ride it out with the ABS, but the longer I braked the more I was losing control...
- What can be done to counteract the rearward bias?? Different pads front and rear?? What do you guys use??

(Or, in translation (and I'm afraid I'm going to be a little mean here, but it's unlikely that this person follows the suspension forum, and it's important to see what the VW engineers have to design for):

"While driving faster than conditions permit and attacking a corner blind, I jumped off the throttle and then hit the brakes in the middle of the corner while my tires were loaded by lateral g's, causing a massive forward weight transfer and generating sudden oversteer. After overcorrecting and almost spinning back the opposite way, I was saved by my rear wheels hitting the curb, before it occurred to me to let up off the brakes. Upon straightening the wheel, I then let the car's natural dynamic stability restore things to equilibrium, after which I congratulated myself on my driving prowess and then blamed the car."

For the people on the suspension forum, it's hard to believe that people can drive this way and expect the car to not spin, but that's exactly what the VW engineers have to cope with. Were our cars a little more neutral, the person making the above post would probably be dead now. And if we add to the scenario a car heavily laden with 900 lbs of family and kid stuff in the back, with random tire pressures and a little bit of winter ice, we can maybe forgive an engineer from thinking, Hmmm, perhaps I'm going to give this car lots and lots of understeer at high cornering loads...

 

[compu_85]

I tell ya what, you lean so much here on TDIClub. I'll defintally be folowing this thread!

-Jason

 

[cartog]

Thanks for this wonderful thread and all the research that lead up to it. I'm particularly appreciative of the background on why things are the way they are.

I've been busy sponging up your threads here and over at vwvortex. Even if I never tweak my suspension, at least I'm learning another dialect of jargon to impress my neighbors with

 

[peter pyce]

Next is what is the Stock Suspension all about The following text belongs to Ceilidh from two years ago. Here the text unchanged:

4a.

Work is getting a little hectic right about now, and I'm finding less time for writing. So let me try something -- rather than write complete chapters (which can take a while to gather & organize thoughts), I'll try writing fragments whenever I have some time (that is, I'll try writing almost in free-form stream of consciousness, instead of working out what to write beforehand). The downside is that things will be harder to read (and much wordier), and the logical flow will be lessened; but the upside is that at least something will be get written down(!). Anyway, I'll try it this evening; if it doesn't work out, I'll go back to the planned out, full chapters.

So here goes for Section 4 sub-a: intro to the stock suspension:
As noted earlier, the stock suspension is (as you'd expect) the most forgiving of the suspension options, and it has the greatest and most persistent understeer. This understeer, however, doesn't quite work the way that many people expect, and if we're going to modify the Golf/Jetta IV suspension for better handling, it'd be useful to know just how and why the stock understeer exists:
Because tuning books and magazine articles spend so much time talking about lateral weight transfer in corners (e.g., stiffening the front of a car makes it understeer; stiffening the rear reduces understeer and leads to oversteer), one can very reasonably get the impression that weight transfer and roll stiffness are the primary determinants of handling balance. And this impression would be reasonably correct if the car in question is the sort of car that tuning books & articles like to talk about: race cars, or street cars that have been heavily tweaked for cornering grip. A hallmark of such cars is good camber control in corners: when a race car or race-modified car goes around a curve, its wheels (or at least the heavily loaded outside wheels) are more or less upright. And with upright wheels, weight transfer is the dominant factor in determining handling balance. But things change quite a bit if, as with the stock Golf/Jetta IV, the wheels are allowed to lean "the wrong way" (i.e., take on adverse camber) in a turn....

Now at this point, everybody's probably saying "Yes, yes, we know all about adverse camber and the stock suspension's stupid inability to keep the tires upright -- that's why we want to modify the car!" -- but here's the kicker: the adverse camber is not "stupid"; it's actually an integral part of what makes the stock suspension so forgiving, and we'll want to be careful about taking it out.

To understand this rather weird-sounding point, let's consider for a moment something very unlike the Golf/Jetta: let's consider a "perfectly balanced" (50:50 weight distribution, AWD, etc.) car with a "perfect" suspension that keeps the wheels perfectly upright in a corner. Such a car will respond "perfectly" to all the well-known suspension tweaks: put stiffer springs or a bigger anti-roll bar at one end of the car, and that end will breakaway first in a corner. Hence if we stiffen the front end a bit, we get mild understeer; stiffen it a lot, and we get heavy understeer, heavy enough to keep even the most ham-handed of drivers from spinning off the road. What could be simpler than that?

But now let's look in more detail at what's going on with the tires: When tires are loaded up in a corner, they initially respond in what's called a "linear" fashion: in colloquial English, they give back exactly what's asked of them -- if you load them a little harder, they corner a little harder; load them a lot harder, and they corner a lot harder; everything stays proportional. This :"linear regime" can persist for some time, but there comes a load after which there's a "transitional phase", where the tires give progressively less and less of what's expected of them, followed by a slip or sliding phase where they can't give any more at all (for the tire aficionados, I realize I'm mixing up concepts from slip angle curves and load curves, so please forgive me! - I'm just trying to get a general qualitative point across...). This passage from linear to transitional to slip is what permits chassis tuning via lateral weight transfer: When we stiffen up the front end of our "perfect" car, we induce understeer by having the front tires (when averaged left & right) hit the transitional and slip phases sooner than do the rears; when the fronts are in transition or slip, they're supplying less cornering grip than are the rears, and thus the front end breaks away first.

All this should sound pretty straightforward and familiar up to this point. But now here's the problem: the transitional and slip phases only set in when the tires are pretty heavily loaded in a turn, and most of the time (when you're on public roads) you're driving in the linear regime --- and in the linear regime, the vast majority of non-race drivers cannot feel whether the car is oversteering, understeering, or driving neutral. Instead, the car simply feels like it is cornering on rails. Moreover, a great majority of non-race drivers cannot discern a tire's entry into the transitional phase, either, and they will only notice something's changed when the tire is well into the slip phase of things. (This is not an indictment of the average driver: it's hard to detect something if you don't know what it feels like, and one of the many benefits of a good performance or race-driving school is that you get practice and instruction in how the tire limits really feel.) Thus for most drivers, our hypothetical "perfect" car will not seem to understeer until front traction is practically used up. Worse still, a tire that's in the slip phase is no longer listening to the steering wheel: once the front tires have entered this part of the cornering curve, the driver is more or less along for the ride. And that's not good.

The above scenario is the origin of the oft-quoted (and oft-misinterpreted) racing adage that "oversteer is where the passenger is scared; understeer is where the driver is scared". When a car with upright tires (e.g., a race car) begins to heavily understeer, the driver can no longer steer. That means that a car set up to moderately or heavily understeer on upright tires is -- for the vast majority of non-race drivers at least -- a very scary beast. Thrown into a curve at high speed, it will seem to corner absolutely neutrally for a very long time, but push too hard, and at some point the front end will "suddenly" and "savagely" break away, at which the steering wheel becomes absolutely useless. That is why race drivers detest heavy understeer; a little is ok, but a lot can be very deadly.

Now, we've just said that this racing adage is oft-misinterpreted, and the misinterpretation comes when we try to apply the adage to street and road cars. Race cars are designed to keep their tires upright in a corner, and understeer is scary. But if you allow a car to roll and to put its tires into adverse camber in a corner, you can transform understeer from a scary, savage beast into something intuitive and utterly benign. As to how that's done...well, I've run out of time, so it'll just have to wait for some other day.....

 

[peter pyce]

And the following, his all the way to the end:

4b.

Right. Now, let's finish up the handling characteristics of the stock suspension.

To review thus far: (1) the stock suspension must understeer enough to keep neophytes from inadvertently spinning under all but the most extreme situations, but (2) creating such understeer via lateral weight transfer & upright tires is unacceptable, as the resulting car will seem (to the untrained driver) to understeer suddenly, savagely, and with almost complete loss of steering control. So what do we do?

We let the car roll, and we let the front tires take on adverse camber while keeping the rear wheels upright.

Let's look at this in detail:

The geometry of the McPherson strut front suspension is discussed fairly exhaustively elsewhere on this forum, so we we'll only briefly summarize the salient points here: when the Golf/Jetta IV rolls on the stock suspension, the outside front wheel (which takes most of the turning load under hard cornering) initially stays fairly upright, but then takes on increasing amounts of adverse camber (i.e., it leans towards the outside of the corner the car is negotiating). This leaning, or adverse cambering, is extremely progressive and continuous: corner lightly, and the tire's pretty upright; corner moderately, and the tire leans moderately; corner hard, and the tire leans a lot.

In contrast, the non-R32 rear suspension is a simple twist-beam with the beam mounted aft of the trailing arm pivot points. Geometrically this suspension behaves much like a semi-trailing arm setup, and the outside rear tire remains relatively upright as the car rolls (at least so long as both rear tires remain on the ground). Thus the geometry of the stock suspension ensures that as the car corners and rolls, the front tires lean (adverse camber) more than the rears, with the amount of leaning progressively increasing with cornering speed.

So why is this good? Ok, now here's the subtle part. When the front tires lean, two things happen: (1) the overall cornering grip at that end of the car goes down, causing understeer in the usual fashion (note that the rear tires stay much more upright); and (2) the tires contribute something called "camber thrust" (for devotees of Milliken and Milliken, yes radials thrust less than do bias plies, but the camber thrust is still there even with radial tires). This camber thrust does wonders for causing progressive understeer in the stock suspension.

Maybe this is a good time to switch to some plain English(!): if you take a tire, and tilt it, it wants to turn in the direction it's tilted; lean it left and the tire will turn left; lean it right and it turns right. This tendency to turn is called "camber thrust". Now, imagine we have a car that is turning to the left: the car rolls to the right, and if the outside front tire is leaning to the right, that tire will want to steer to the right as well. This steering to the right causes understeer (because the driver wants to turn to the left), but -- and here is the important point -- the understeer has nothing to do with the tire nearing its traction/grip limits. In fact, the understeer will occur quite vividly even when the tire is in its "linear" regime, where it still has all the traction in the world. This is a very, very useful phenomenon, because it means that when the car starts to understeer, the driver can still steer the car. In contrast to the upright-tire (race car) situation where significant understeer occurs only after the front tires have used up much of their traction (at which point they stop responding to steering inputs), the leaning-tire understeer occurs early enough that steering is still an option, and the amount of understeer progressively increases with cornering load.

Because the front tires on a stock Golf/Jetta lean, the car behaves in a very "natural" fashion as it corners: take a corner mildly, and the car simply goes around unfussed; go a little faster, and it understeers a little; go moderately faster and the understeer becomes moderate; go way too fast and the understeer becomes pretty heavy. But at no time does the front suddenly and savagely break away, leaving the driver helpless to do anything but stare at whatever he's about to hit. At all times, the driver can correct for the understeer simply by turning the steering wheel a little harder: because the front tires are still well below their traction limits (despite all the understeer), they will still steer the car back onto the cornering line. Thus whereas with a race car, "understeer is where the driver is scared", with the stock car understeer is not scary at all -- it's just slow and rather annoying.

Bottom line, the roll-induced moderate understeer on the stock Golf/Jetta IV provides some very nice characteristics for driving on public roads: if you go too fast, the front of the car gradually begins to drift wide of the intended line, and you can correct your path simply by slowing down or by turning the steering wheel a little or a lot harder. These handling characteristics are so "natural"-feeling that most drivers assume they're some sort of law of nature. They're not. They're there because of moderate understeer, induced by roll, and they're not present in a true race car.

(By the way, the stock car is set up so that the roll-induced understeer becomes intolerably unpleasant long before the tires hit their traction limits; on dry roads, it's very difficult for any but the truly masochistic to get to the point that the front tires no longer steer. If you do get to that point (it's more likely to occur on the slower corners), you'll feel the car just start "plowing forward" with a lot of tire scrubbing noises....)

Digression: a Pathological "Tuned" VW

There's still a bit more to be said about the stock suspension, but anyone wading through these last two "stock suspension" installments will probably be wondering why we're going through all this in a vortex thread that's ostensibly about improving handling. So I'll give a brief taster here, as an example of why an understanding of the stock suspension can help us in modifying our cars.

We'll consider a pathological case:

A couple of months ago, there was an interesting thread about anti-roll bars on this forum, and one of the individuals posting appeared to be a pretty nice, perceptive fellow who had been given some very bad advice. This advice had led him to install a fairly gargantuan front anti-roll bar on a car that was in other respects not far from stock. His report was that even with the huge front bar, understeer was significantly reduced and the handling was much sharper -- but in wet weather, the front would sometimes break away very suddenly and unexpectedly. Let's look at why this should happen:

First of all, remember that when upright tires are working in their linear regime, it's extremely difficult for untrained (non-race) drivers to discern whether a car is understeering, oversteering, or neutral -- the car simply feels like it's on rails. Secondly, remember that the stock suspension, with its leaning tires, is designed to give camber-thrust-induced understeer while the tires are still well within the linear regime. Put these two concepts together, and....

If you take a stock car and reduce the amount of roll, there is less tire lean, less camber-thrust, and less understeer -- almost irrespective of how you go about reducing the roll (at least so long as you're in the linear regime). Hence if you mount a gargantuan front bar on an otherwise stock car, it understeers less simply because it rolls less -- so long as you're going at moderate speeds. A trained driver, one who's used to driving racing-style cars, will of course detect almost immediately that the car is still understeering (because of the greatly increased front lateral weight transfer, caused by the massive bar), but an untrained driver will only note the cornering-on-rails sensation of upright tires in the linear regime, and he'll think that the understeer's gone. Hence the report that "Mounting a big front antiroll bar eliminated the stock understeer on my car!".

But now let's look at what happens in the rain: on dry roads, the cornering limits of a flat-cornering VW are high enough that most neophytes will not reach them; hence the driver of the big-front-bar car probably never left the linear regime, and so he never felt the car understeer. But in the wet, the tire limits are much lower, and they can be reached when the bar is still holding the car very flat (with the tires still upright). As a consequence, the big-front-bar car in the rain handles very much like an upright-tire race car with far, far too much weight-transfer-induced understeer: when corners are taken too fast, the fronts exceed their traction limits, and they break away in the regime where they no longer steer. Thus the supposedly understeer-free car in the dry takes on a sudden, frightening understeer in the wet. Not good.

Pathological Example #2

I wasn't going to do this (we're getting a little ahead of ourselves), but as long as we're talking about pathological suspension setups, let's also look at one of the very many reasons why a Golf/Jetta IV with a massive rear bar and stock everything else is also a very ill-handling beast (or, as Dick Shine has often warned people, a rear anti-roll bar cannot cure a fundamentally unsound suspension!).

We'll talk about the big-rear-bar/ soft front setups more in a later installment, but for now we'll just focus on the very common thread postings where someone (usually a novice) raves about the 28mm bar on the rear of his stock car, but then warns someone else (also usually a novice) to take the bar off in the winter, because otherwise there's sudden savage oversteer (usually these email exchanges have horrible misspellings, by the way, but that's a different topic...). Leaving aside the questionable merits of a suspension you have to remove on a seasonal basis, what's really going on here, and is this a good setup?

1) One possibility is that the novices never noticed any oversteer in dry weather because they never pushed hard enough to leave the linear regime. Again, when you make a car roll less than stock (here by installing a big rear bar), you make the front tires more upright, the roll-induced understeer decreases, and so long as you're in the linear regime, everything feels great. And also again, the actual performance limits of even the stock car (which invariably feels much closer to the limit than it actually is) are high enough that many drivers don't approach them in the dry. So one possibility is that the car was an oversteerer in both summer and winter, but only in winter was it noticeable. Perhaps.

2) But let's look at the dry, summer handling a little more closely. For reasons we'll perhaps get to later, someday, a big front bar can substantially reduce roll, but a big rear bar will only reduce the initial amount of roll. At some point, the inside rear wheel will lift (it'll lift on even a well-driven stock car), and when that happens, it really doesn't matter whether there's a big anti-roll bar in the back or not: as far as the car is concerned, there is one wheel on the ground in the back, and two wheels on stock springs and stock bar in the front. Hence at the dry-road 3-wheel cornering limit, a big-rear-bar car will understeer about as much as will a completely stock car (note: there's actually a bit of a difference in that a big rear bar will, by lifting the inside wheel higher, cause more leaning of the outside wheel, but there's so much understeer built into the stock suspension that the overall result is still understeer). Now, this is a fairly horrible handling set up (understeer sets in rapidly the moment the rear wheel leaves the ground), but there's no oversteer at the limit.

But what happens in snow? When the roads are truly slick with snow or ice, the tires reach their performance limit at a very low g-force -- so low that the car has hardly rolled at all. Thus in the snow, the big-rear-bar car (with stock front) behaves like an upright-tire race car with far, far too much lateral weight transfer at the rear: when the limits are reached, the rear breaks away, suddenly and savagely. Thus you have a car that seems neutral at low to moderate speeds in the dry, that understeers heavily at the dry limit, but which oversteers dramatically on slick roads.

And for this someone has paid $300?

 

[Golf_GTDI]

I think this is where we sprinkle in some of Peters famous suspension animation to help illustrate.

 

[bhtooefr]

My car has a bit of a darter feel to it, FWIW... then again, it IS a Mk2...

Thread WATCHED.

 

[4vdubs]

I'm waiting with bated breath for the next installment! How do we make a MkIV better?

 

[bhtooefr]

4vdubs: What kind of handling are you targeting? What's your definition of better?

FWIW, if you're targeting "darter" handling, step one is to do an oil cap mod on your Mk4. Take the oil cap off, sell the Mk4, get a Mk2, put the oil cap on.

 

[4vdubs]

4vdubs: What kind of handling are you targeting? What's your definition of better?

FWIW, if you're targeting "darter" handling, step one is to do an oil cap mod on your Mk4. Take the oil cap off, sell the Mk4, get a Mk2, put the oil cap on.

Lol...I'm keeping mine thanks! I'm looking for the best street handling (grip) I can get out of my Mk4, with a little autocross thrown in to keep life interesting. I know this thing is far from perfect but I plan to make the best of it. I have a shine RSB and everything else is stock for now. With the Shine bar the car seems very neutral at least on dry roads as Peter says. I have explore the limits somewhat on some smooth dirt roads and experienced the oversteer but it seems easily controlable with a little counter steering. I await further advice from Professor Pyce.

Scott

 

[blacka5]

Hello Everyone,
<understeer ...>

OK, color me confused. I like to think of it the following way (and where I'm wrong please correct)

Tires are not rigid objects, so in a corner the wheels and tires don't point in exactly the same direction. The angle formed by the crown line of the wheel and that of the tire at the contact patch is the slip angle, which is zero when the car is going in a straight line. In a corner, at any given instant, the crown lines of the front and rear tires will point in the same direction, but it doesn't follow that the crown lines of the wheels do. If they do, then the front and rear slip angles are equal, and it's a neutral steering condition. If the front SA > rear SA, that's understeer, and if the front SA < rear SA, that's oversteer.

On applying power in a front driver in a curve, you are asking the tire to accept a power input as well as a steering input, so the tire deforms more and the front SA increases, leading to understeer. You can induce a transient oversteer condition by lifting or braking, which unweights the back end and allows it to swing out, but you can't maintain oversteer in a front driver, not without pathological suspension geometry, anyway. On a rear-driver, OTOH, enough throttle input can cause the rear SA to increase to a value greater than the front, so you can maintain a power oversteer condition. You can, of course generate a transient oversteer in a rear-driver by lifting off, especially if you have swing axles (a la Triumph Spitfire) or semitrailing arms (especially^2 if you stick something heavy way out back, a la Porsche 911), although in the latter case, it's really an effective SA, since the geometry points the rear wheel outwards on roll loading. Other things being equal, a rear-driver will handle better than a front because 1) the rear wheels do real work, not just keep your hind end from dragging on the road, and the front only needs to deal with steering inputs; and 2) you've decoupled steering and power inputs, so you can use both ends of the car to adjust your attitude in the turn.

That's how I understand it. I do like the specific details re:suspension geometry, tire loading, etc -- please keep it up!

 

[Nate_Grauvogel]

Thanks, Peter & Ceilidh, for posting the fruits of your hard-won experience here in this more orderly venue... I go to the Vortex to look at the classifieds by those getting rid of stock pieces, but not much else. Don't have the patience.

I understand all you've said so far, and it jives with what I've learned in theory and simulation in my Vehicle Dynamics class (after May graduation I'll have time to delve deeper in that subject on my own... not now though)... we focused on a racier setup so we wouldn't have to factor in changing camber effects on tire performance... the modeling of such was beyond the allotted semester time. Anyhow - thanks for sharing... if people will stop and listen, you will make a lot more people informed if they wish to be. I like your categories, BTW.

In DFW, I should be after Slow Car Fast, but I used to live in the NM mountains, where GT would have been the sure ticket to wheeled nirvana... hope to get back there someday.

Keep preaching!

 

[peter pyce]

I am sure he will come soon and address some of the questions, especially those regarding the Golf A2 and the Spitfire, as he actually used to own both of these cars, so stay tuned. Just give him time as the man is very busy with more important things in life.

Meanwhile, let's continue with his outstanding poetry on handling. Follows a copy-paste from the next chapter - the autocross setup. The following all his to the end:


5. Autocross Suspensions

There are a few more things to mention with the stock suspension, but we can cover those points when we discuss GT setups. For now, let's look at autocross.

This will be comparatively brief, as I (Ceilidh) have never autocrossed or set up a car for autocross. Still, a brief pass through the theory will make the GT setups (our ultimate destination) more intuitive.

So here we go: As discussed earlier, autocrossers drive at low speeds on smooth surfaces, and spend most of their time timing and transitioning between turns. Agility is at a premium, as is cornering grip, and (as we're talking track cars here) we don't care about ride quality or progressive, forgiving understeer, and we want the car to be neutral. Autocross cars make perhaps the most use of the various go-fast products on sale for VW use, and there's more room for variety here than anywhere else. Here, in terms of general theory, is why:

I. A Hypothetical, Perfectly Smooth Track

Imagine we're running on a perfectly smooth autocross track, one where there is not even the slightest suggestion of a bump, crack, or pavement ripple (this is the assumption that most tuning magazine articles seem to make, by the way, so what follows next will probably sound familiar). As we don't have to worry at all about road imperfections (in this hypothetical, completely unrealistic scenario), the basic setup theory is straightforward:

A) The more upright the tires are (especially the outside tires in a corner), the more cornering grip we can get.
B) The more we can get the car's weight evenly distributed among all four tires, the more ultimate grip we have.
C) The less the car rolls, and the stiffer the "roll rate", the less time the car spends taking a set for a corner, and the more rapidly we can transition from corner to corner.

To achieve these goals, we can "tune" our car exactly the way the tuner mags recommend:

1) We install massive springs and/or antiroll bars to cut roll to a minimum, and to increase the roll rate (the "springiness" of the car in roll). This modification keeps the front tires far more upright in a corner (which increases grip and cuts understeer) and speeds the car through transitions (because the car takes a set much more rapidly on corner entry).

2) We force the wheels into negative camber so as to counteract any residual roll, and to perhaps even get camber thrust to work with us by leaning the outside tires towards the inside of a corner.

3) We balance the springs/bars fore & aft so that the cornering balance is neutral (as the car is nose heavy, that means the rear should be a bit stiffer).

4) And we drop the car, front and back, as far as we possibly can, so as to lower the center of gravity (CG). Lowering the CG does not reduce roll -- it in fact increases it because of the oft-discussed rapid lowering of the front roll center, but we can counteract the increased rolling tendency simply by making the springs and bars still stiffer. The lowered CD does, however, reduce the total lateral weight transfer in a corner (the total transfer is simply a function of track width and CG height, and is independent of suspension design or spring/bar rates). This reduction in lateral weight transfer means the 4 tires are more evenly loaded in a corner, which increases overall grip and cornering speeds.

4-A) As a side note, dropping the car leads to a very low front roll center (it's actually below the ground now). For reasons we won't go into here (it's discussed in all the good textbooks), the roll center height controls lateral weight transfer at the initial instant of corner entry, and a lowered front roll center thus means better initial turn in.

5) We install big, stiff, quick-acting Bilstein or other monotube shocks to further snub down the chassis and thus speed up our speed in transitions.

6) And to further reduce play or compliance in the suspension (all of which increases the time it takes for the car to take a set, and/or reduces the effectiveness of our antiroll bars), we replace all the bushings with spherical joints, solid bushings, etc., etc.

The above really should sound extremely familiar, as it's what all the tuning shops and tuning magazines tell us to do.

So what's the problem? Well, for one thing, even an autocross track isn't perfectly smooth. And that has the potential to change everything:

II. The Problem With Bumps

An implicit assumption in the "perfect world" autocross setup is that we can make the springs and bars (and bushings, shocks, etc.) as stiff as we need to in order to control roll and body sway/heave/roll/pitch/etc. But in practice, a car's suspension exists for a reason, even in a race car where ride comfort is a non-issue: When there are bumps, a perfectly rigid car with a perfectly rigid suspension will spend most of its time with its tires flying from bump crest to bump crest, and during the time the tires are flying, the cornering grip is zero. The suspension on even a track car thus has the critical job of keeping the tires pressed to the pavement as evenly and as firmly as possible, and that requirement puts a limit on how stiff we can set the springs. Thus we have to add two new goals to A, B, C listed above:

D. The softer the springs are and the better tuned the shocks are to the spring rates, the better the "mechanical grip" between tire and track surface.

E. The more independent the suspension truly is, the better the mechanical grip.

Requirements "D" and "E" go at odds with many of the "standard" tuning tweaks advocated by the go-fast vendors. The complications are many, but here are some of the biggies:

1) If we stiffen the springs too hard, the tires lose mechanical grip. This puts a limit to how stiff we can go in an effort to control roll.

2) If we try to circumvent #1 by increasing the antiroll bar size, we eventually begin to lose grip by (a) non-independent suspension action (the bars tie the inside wheels to the outside, and start behaving like a solid axle suspension), and (b) an inability to tune the shock absorbers to match both the (softer) vertical spring rate and the (stiffer) roll rate.

3) Lowering the front suspension rapidly drops the front roll center, which increases the roll couple; the car therefore either rolls more, or requires stiffer springs to counteract the roll (which reduces mechanical grip). Both of these effects negate or partially negate the lateral weight-transfer benefits of lowering.

Thus the appearance of requirements "D" and "E" mean we can't simply slam the Golf/Jetta IV down to the pavement and stiffen the springs. From the autocross link that Alexb75 posted a couple of weeks ago, it appears that at least some autocrossers are actively experimenting with different avenues of optimization. In general theory, at least, one can try to:

1) go with a lowered car, and back off on the springs & bars (thereby rolling more, but maintaining mechanical grip and reduced lateral weight transfer)

2) keep the front end relatively high, and fit moderate springs & bars (which rolls less, keeping the tires upright and maintaining mechanical grip, but accepting more lateral weight transfer)

3) go moderate on the springs, but go stiff with the bars (which maintains mechanical grip on vertical bumps, but loses out on 1-wheel bumps, and in some cases loses out on transitions because of play in the antiroll bar linkage)

4) stay moderate on the bars, but go heavy with the springs (which loses general mechanical grip, but improves agility (non-Shine-style bars typically have a bit of play, whereas springs act right away in transitions)

5) or try various combinations of 1, 2, 3, or 4, and try different things on different ends of the car (e.g., soft rear springs and big Shine bar, plus stiff front springs and no bar, etc.).

What will actually work best? That's a question for the autocross experts, not me (Ceilidh)! But the main points here are that:

A) with Autocross setups, different combinations might work best on different tracks (e.g., a smooth tight track, vs. a rough, faster track, etc.) and with different drivers .

B) But far more importantly for the purposes of this thread, the need to absorb bumps is something that just can't be ignored in setting up a Golf/Jetta IV suspension -- even on something as comparatively smooth as an autocross track. If bumps didn't exist, we'd just mindlessly set up our cars the way the speed magazine breathlessly extoll (slam the car, stiffen the springs, mount big bars front & rear, etc.). But when reality sets in, in the form of actual track and road surfaces, the ideal speed-magazine setup gradually morphs into something a lot closer to the Shine -- and when we factor in ride comfort later on in this thread, the Shine will start to look even better.

Next up: road racing theory

 

[peter pyce]

Let's continue with the copy-paste. Ceilidh on Road Racing and Shine all the way to the end....

6. Road Racing, and the Shine Real Street Suspension

Let’s take a brief look at the general characteristics of a decent road-race setup, after which we'll touch for a moment on a setup that gets a lot of press in these pages: the Shine SRSS.

Backdrop

If you can remember back to the Autocross installment, we started there by considering what the ideal setup would be if we ran on perfectly smooth roads (e.g., lower the car to drop the CG, stiffen the springs and bars hugely to reduce roll, go to negative camber to keep the outside tires upright, install ultrastiff shocks to control body movements during transitions, etc.); then we looked at how the unavoidable presence of bumps -- even on a smooth autocross track -- forces us to dial back on those "ideal" suspension mods; and finally we concluded by observing that, for autocross at least, the jury's still out on what's the ideal compromise....

If there's room for Golf/Jetta IV variety in Autocross, it'll be because of a tradeoff between the simplistic benefits of lowering & stiffening vs. the more subtle ability to maintain traction and control on bumps and rough surfaces. Because Autocross surfaces are in fact fairly smooth, and because the low vehicle speeds in Autocrossing make the bumps seem even smaller, it's conceivable that a low, very stiff setup might actually work (we'll leave it to experts to comment on whether such setups do or do not work in reality -- all we're saying here is that if such setups are to work at all, it'd be in Autocross...).

When we get to road racing, however, the much higher speeds and more varied terrain mean we can no longer even pretend to ignore the bumps, and that seems to pretty much rule out the ultrastiff, very low setups for the Golf/Jetta IV. In the Autocross installment, we touched on some of the reasons why that might be so, but let's now look at it with a little more detail:

Spring Rates and Mechanical / Rough Road Grip

As discussed earlier, for every combination of road surface, vehicle weight, unsprung weight, tire stiffness, and vehicle speed, there is a certain suspension spring rate (and shock damping) that will maximize tire grip. If the spring is too stiff, the vehicle is thrown into the air with every bump; if the spring is too soft, the vehicle might ride smoothly, but the inertia of the unsprung weight will carry the tires upwards and off the ground on each bump, and the springs will be too weak to force them back down. Both situations will lead to the tires unloading on the back side of bumps (in the too-stiff case, because the whole vehicle is rigidly leaping about; in the too-soft case, because the tires alone are hopping up and down), and traction is therefore lost.

(As an aside, the above is one of the several reasons why traditional sport utilities & trucks ride and handle so badly on rutted pavement: an old truck-based SUV has very heavy solid axles with a lot of inertia, and that causes the axles to leap about on rough roads. Soften the springs for a better ride, and the axles bounce around almost uncontrollably; stiffen the springs to force the tires back onto the pavement, and the whole truck gets jolted with every bump. It's partly for this reason that almost all the new generation SUVs are going over to independent suspension: one advantage of independent suspension (on a driven axle) is drastically reduced unsprung weight...)

So, in theory one has to carefully choose the spring rate on a car to maximize traction on a real road or track, as too-stiff and too-soft are both problematic. But in practice, at least with Golf/Jetta IV's, we don't have to worry about too-soft. Our independently-suspended cars are heavy, and the CG (even on a slammed car) is high relative to the track and wheelbase; hence if we start softening the springs, we run into problems with body roll, dive, squat, heave, etc. long before the unsprung weight gets out of control.

Or to put it another way: to control the body motions on our cars, we have to stiffen the springs enough that they are almost always too stiff for optimal mechanical grip. Hence, in practical terms, we have a more or less clear-cut tradeoff: stiffen the springs to control roll (and pitch, heave, etc.), or soften them to improve the tires' grip on the track.

Implications of Reduced Spring Rate -- Why Lowering Causes Problems

Thus because of bumps, we can't arbitrarily stiffen the springs to control roll. (And remember from the earliest installments: if the car rolls, the stock Golf/Jetta suspension is intentionally designed so as to progressively lean the outside front tire, which reduces grip and causes moderate to heavy understeer.) So once we leave the smooth, low speeds of the autocross track (and some would argue even *on* the smooth, slow autocross tracks), we have to soften the springs. And that softening causes a whole litany of problems for a lowered car, of which we'll highlight two of the biggest:

1) When we earlier discussed the "ideal" setup on a "perfect", smooth autocross track, we rather naively imagined that we could stiffen the springs to the point that the chassis no longer moved: our "perfect" setup wouldn't roll in corners, and neither would it pitch, dive, or squat under acceleration or braking. Whether or not that's a reasonable approximation on an autocross track is something we'll let the autocrossers debate, but once we're on a real road-race track, with real bumps and realistic spring rates, we can no longer pretend the body isn't moving around. It will move relative to the wheels as the car accelerates, decelerates, corners, and encounters bumps. And that movement has some big geometric implications, particularly at the front suspension:

For a variety of geometric and packaging reasons, it's almost impossible to design a production car front suspension that doesn't "bump steer": when the front wheels move up and down relative to the chassis, they don't stay pointing straight ahead -- instead, at some point they will begin to toe-in or toe-out.

Such toeing would be pretty undesirable for fairly obvious reasons: it means that the toe setting changes when the car pitches forward under braking, or when it pitches back under acceleration, or when it rises and falls over bumps. Even worse, if the car rolls, one tire might be toeing in while the other is toeing out (because one wheel is rising while the other is falling), which steers the car to one side (called "roll steer"); or if one wheel hits a bump but the other doesn't, the bumped wheel can toe in or out while the other keeps pointing straight, which again steers the car (called "bump steer"). Mix and match these various situations -- e.g., let's simultaneously decelerate and roll the car via trail-braking, and then hit a one-wheel bump on corner entry -- and the combined roll and bump steer effects can be extremely entertaining and ever-changing, thereby inspiring the driver to generously & politely compliment the race engineer for the wonderful setup (drivers really love race cars that dart about unpredictably in corners)....

As even production car suspension designers don't like to be shouted at, a good deal of engineering time is spent making sure that bump steer is rarely an issue in the normal life of a car. In practice that means specifying a geometry that crams all the toeing to the far limits of the suspension motions: so long as the wheels are moving up and down to positions reasonably close to the static load position, the toeing is negligible; only when the wheels move close to full jounce (all the way up) or full rebound (or all the way down) does the toeing become pronounced (if you look at a graph of toe vs. suspension movement, the graph is often a straight line (essentially zero toeing) for a good distance above and below the static load position, but then takes a pronounced hook as it approaches full jounce or full rebound). In this way bump & roll steer are ordinarily non-issues: under normal braking, accelerating, and cornering, the car handles fine, and bumpsteer only shows up on bumps so enormous that any steering effect is swamped out by all the other violent things that must be simultaneously going on.

So what happens when you drastically lower a production car? In severe cases, you move the static load position (the position where the suspension sits and works around) right into the region that contains all the bump steer. And so the car bump steers. It also roll steers. And the turn-in characteristics (which are greatly affected by toe in and toe out) will vary wildly depending on how hard you brake during corner entry, or whether you're hitting 2-wheel bumps. In short, your "race car" will adopt many of the delightful handling characteristics of a 1940's Buick, minus the comfortable ride and cool hubcaps.

(2 notes here: one is that the rear suspension will bump & roll steer as well, though the effect is usually less pronounced than in the front; the other is that it's standard practice to "bump steer" (meaning, to "reduce the bump steer effect on") a radically lowered car by changing the positions of the steering rack and steering arm pickup points; but I've never seen any discussion of such on the VW forums -- probably because the next point (below) makes the issue moot)

2) The second effect is one that's been discussed to death elsewhere on this forum, and which we'll only repeat here so as to put in the context of the earlier installments: Almost everyone following this thread will have already read somewhere that lowering a McPherson strut suspension will increase the tendency to roll. Some readers, however, might have wondered why that's an issue: if the car tends to roll more, why can't we just stiffen the springs to compensate?

The reason, as it should be clear by now, is that we can't arbitrarily stiffen the springs. If we stiffen everything so as to reduce roll, we lose mechanical grip over realistic bumps; if we soften the springs to increase mechanical grip, we roll the outside front tire into adverse camber. Catch-22. What we need, then, is a way to reduce roll without going crazy on the spring rates. And we can do that by keeping the roll center nice and high.

Such the reason for Dick Shine's oft-repeated assertion (oft-repeated not because he's been unclear, but because so many people seem unwilling to believe him) that lowering the front end of a Golf/Jetta IV will destroy the handling. If you don't keep the roll center fairly high, you lose grip either through increased roll & adverse cambering of the outside tire, or else through having to control the roll with overly stiffened springs. (Note: we'll not discuss drop spindles and major component swaps here).

Or to summarize it another way: there are three big means of increasing/ maintaining the cornering ability on our cars: (1) keeping the tires more upright by reducing roll; (2) retaining mechanical grip by not letting the spring rates get out of hand; and (3) reducing the total lateral weight transfer by lowering the CG. On a Golf/Jetta IV, #1 is the most important, and #3 is the least. Hence optimizing #3 while compromising #1 and/or #2 is not the way to go.

Continues below ……

 

[peter pyce]

Continues from above on the Road Racing and Shine SRSS .... (Ceilidh):

Back again. =) In re-reading the earlier post, I see that somewhere along the way I morphed from true "road-race" to "streetable semi-track" setups; sorry about that. Most of what we're talking about now concerns cars that live somewhere between road and track, and our ultimate destination (given the original topic of this thread) is the GT suspension. for the cars we're talking about here, rough road performance is becoming increasingly important.

Right. Where were we? At the end of the last installment, we reviewed the oft-discussed reasons for why we have to keep the front end high. Now we'll clear up some additional points, and then go into just how high is high enough:


Caveats and Additional Notes

A) Before I forget: road racers tend to go for negative camber, which is beneficial on the track for a number of reasons (see Carroll Smith's "Tune to Win" for a discussion). As a personal caveat (from my experiences with "bigger" negative camber on RWD cars): if you overdo it, you can mess up your braking (the tires no longer sit flat even when the car's on the straight, and brake-dive makes the cambering even worse); you might wear your tires pretty unevenly (the inside edges go first, if you're not regularly cornering hard); you can lose on-center feel and a nice street-style progression on gentle curves (the steering becomes 2-step: first you're riding on the inside tire edges, then at some point in a corner the outside tire flops down flat and begins rolling onto the outside edge); and sometimes you get tramlining issues.

B) On a different note: some of you might be wondering "Why does Ceilidh keep going on about stiff springs? If the problem concerns roll vs. mechanical grip, why not use soft springs for grip, and reduce roll with antiroll bars?". That's a fine question, which we'll go more into later when we get to GT setups, but for now please just consider that a stiff bar (1) increases the 1-wheel bump rate; (2) reduces the independent action of a supposedly "independent suspension"; and (3) makes it hard to tune the shock damping to simultaneously suit 1-wheel and 2-wheel bumps. All these factors reduce mechanical grip, and thus there's a practical limit to how stiff the bar can be. In short, bars are nice, and they allow you to control roll with softer springs (compared to what you need if you didn't have bars at all), but ultimately you still have to worry about spring rate.

C) Last note: One thing I've completely ignored up to now is the degree of camber gain under roll (which determines how upright the tires stay as the car rolls). It's an important concept, but I've left it out because it's so straightforward, and because it's been so well explained on other threads. But for completeness: a McPherson strut suspension that's rolling will initially tend to keep the outside tire more or less upright; but with increasing roll, the outside tire will begin to lean (and thus incur camber thrust, reduced traction, etc.). When a strut suspension is dropped via lowering springs, the effect on the outside tire is curious: at small angles of vehicle roll, the tire is more upright than before; but at greater roll angles, the tire actually leans more than in the stock case. Hence if you try to lower the CG by installing lowering springs, and then try to maintain mechanical grip by keeping the springs relatively soft, not only do you get more roll, but the adverse camber on the critically-important outside front tire is even worse than you'd first expect.


How High Should the Front Suspension Ride?

Everything up to now has been qualitative: we've discussed why a lowish front roll center can be bad; how excessive lowering can cause bumpsteer problems; and how there are camber-gain issues if we lower too much and allow too much roll. But does that mean we can't lower at all, and should we be pushing things much higher than stock?

The second question is more easily dealt with: One might ask -- given that a stock front ride height is good -- whether a much-higher-than-stock front would be even better. The short answer is "No": aside from the bump steer problems you'd get from pushing the car towards the extremes of rebound (vs. jounce, with lowering springs), at some point the increased lateral weight transfer you get from a high CG (plus the positive static camber you get from a raised McPherson geometry) will hurt you more than you'll benefit from reducing roll and/or spring rate. So it's very possible to go too high.

The other question is less easy to answer: just where should the front end sit? The roll-center and camber-gain phenomena are not "step functions" -- it's not a case of, go 1mm too low and WHAM! the car is screwed up. And bumpsteer should not be an issue (on any modern production car) if you're within an inch or so of stock height, at the very least. So maybe we can go down an inch, or a half inch, or somewhere in between? In short:

"How low is too low?", and "How can one tell what's best?".

The answer to the first question is "It depends.", and the response to the second is that, unless you want to exhaustively test lots of different setups, "You'll just have to trust someone".

Let's explain:

For every car, there will be a magic combination of ride height and spring rate that will optimize the combination of (1) reduction in roll; (2) maximization of mechanical grip; (3) best use of camber gain in roll; and (4) minimization of total lateral weight transfer. That magic combination does not, a priori, have to be at stock height, and it will not be the same for all cars -- not even all cars that use a McPherson front suspension. For example, some people have suggested that BMW's seem to work just fine with moderate lowering; well, that's entirely possible. But as BMW's have a different CG height, track width, tire & rim size, camber gain, suspension travel, etc. from our VW's, their ability to work with lowering tells us nothing about what will work best with the Golf/Jetta IV.

Since this particular issue comes up so often on various threads, it might be worth repeating: just because race winning Beemers, or Hondas, or Toyotas, or etc. happen to use a particular ride height, doesn't mean that a VW should be set up the same way. Every car model is unique, and the optimum chassis setup is similarly unique.

So while I'm afraid it'll sound like a major cop-out, if we want to know what the ideal ride height is for a Golf/Jetta IV, and if we don't want to expend vast sums of $ and time to experiment with different combinations, we pretty much have to rely on the judgement of people who (1) have themselves performed lots of experiments; who (2) possess the experience and the know-how to assess what those experiments are telling them; and who (3) don't have an enormous amount to gain by telling us something that's not true.

To summarize, repeat, and stress once again: there's nothing inherent about a McPherson strut front suspension that requires the ideal performance variant to sit 0.5" above stock -- but neither is there any reason for it to be at any other height, whether higher or lower. What determines the ideal height is the particular combination of parameters (CG location, track & wheelbase, camber gain, etc.) designed into a particular car, and we as ordinary consumers have to trust somebody who has exhaustively tested different combinations to determine what is best. I (Ceilidh) would tend to trust Dick Shine, but you (dear reader) will have to decide for yourselves whom you will listen to! But regardless of whom you choose, please pick your "expert" on the basis of what he/she says concerning issues other than ride height (that is, how credible is this person on other issues?) -- regarding the Golf/Jetta IV front ride height, there is no a priori reason why it should be high, low, or anything in between.


The Shine Setup

And finally (and this will have to be brief, as I'm late for a meeting! ) -- if we take as a given that the Shine SRSS works well as a mild track setup, here's a possible theoretical explanation for it (Dick Shine, please do chime in and correct me if I go astray here!).

From the perspective of an outside armchair observer, the Shine philosophy seems to be to:

A) keep the front at or just above stock height to minimize roll with the use of reasonable spring rates;
B) increase front spring rates to control roll and pitch, but to keep the rates moderate, so as to retain maximum front mechanical grip;
C) refrain from increasing the front bar beyond stock size, again so as to maximize front mechanical grip (as briefly explained above);
D) increase rear spring rate to the limits of good mechanical grip and a tolerable ride while maintaining good balance for at-the-limit cornering;
E) drop the rear ride height so as to lower the overall CG (the rear roll center drops slightly less rapidly than does the CG, so the overall roll couple is also somewhat reduced); (this lowering of the rear also allows for more stiffening of the springs & bars, as it helps keep the car from getting too tail-happy)
F) provide the option of a rear bar to adjust handling balance, particularly in transitions. For reasons we'll discuss in a future installment, the rear bar does not significantly affect the ultimate grip of the Shine suspension (when the car is three-wheeling), but it sharpens turn-in by keeping the car flatter (on turn-in and transitions) and by increasing lateral weight transfer at the rear (again, primarily on turn-in and transitions). To have a noticeable effect on what is already a moderately stiffened suspension, the rear bar has to be fairly stiff; this stiff bar reduces mechanical grip in the rear for all the reasons alluded to earlier, but as the purpose of the bar is to reduce understeer in the appropriate situations, this loss of grip is acceptable.

 

[Ceilidh]

Regrouping a little...

Hello Folks,

First of all, an apology for not replying to all the kind folks who have posted comments; life is a bit busy these days, but let me try to go through the posts, in reverse chronological order (newest first):

1. Nate, that sounds like a very interesting course that you're taking -- If you come up with any neat insights, do please tell everyone! (What software package are you using? Would you recommend it?). And where in NM did you live? I did the drive up from Los Alamos to Mesa Verde via Chama a few times: what heavenly roads!! Hope you get back out there sometime...

2. blacka5, your understanding is perfectly correct, and very well explained! The traction effects you discuss do in fact explain how you can get steady-state under- or over-steer in a theoretically neutral vehicle, and as you point out, it's pretty hard to induce continuous oversteer in a FWD car (I suppose you could do it via forward weight transfer on a continuous downhill curve, but that'd require you to drive down a spiral parking lot ramp, or something like that!).

To set blacka5's comments in a larger context: there's a teaching aid called the "friction circle" (Peter, can we have a picture?), which is used to show that a tire forced to exert acceleration or braking forces has less traction available for cornering (the idea is that when you're at 45 degrees on the circle, half of the traction is used for cornering, and half for braking, etc.). So long as the size of the circles (which represents the total traction available at each tire) remains fixed, spooling off some of the traction for acceleration will cause a RWD car to shift towards more oversteer (or less understeer), and a FWD car to shift towards more understeer.

(As an aside, this effect used to be much more vivid in the old days, when tires were skinny and slippery: I had an MGB GT that was just a joy to drive slowly on winding roads, as the engine torque needed just to maintain a constant speed was enough to drift the (RWD) back end outwards, and long sweepers could literally be driven on the throttle (a little more gas to tighten the turn, a little less to open it up -- essentially the opposite of how our front-drivers work in practice...)

To finish up the context: the points raised by blacka5 apply when the friction circles remain constant in size, but they're complicated when the circles change in size with different situations. With all cars, the rear circles grow and the front circles shrink with acceleration (because weight is shifted rearwards, increasing the rear traction while diminishing the front), whilst the opposite happens under braking. This growing & shrinking can sometimes overwhelm the "normal" friction circle effects (e.g., on that same MGB GT, the car was far more stable (less prone to spinning) in tight corners when under power, as the acceleration would increase the rear traction and foster some stabilizing understeer).

(Digressionary aside #2: one of the joys of those old British sports cars was how the handling balance would shift with circumstance. In some situations, more throttle would give you understeer, whilst in others it would give you oversteer -- sometimes the trend would even change in the course of a single corner! It sounds like a fault, but it wasn't! The cars simply gave you a really vivid idea of what was happening in terms of physics, and you could learn an awful lot without even breaking the speed limit; it was fun!)

(Digressionary aside #3: this has nothing to do with anything, but at one point I had my college roomate at the wheel of my car, and he kept scaring both of us by repeatedly slowing down at every corner -- each time he did, the car would get squirrely, which would scare him more, causing him to slow down even more on the next corner, which would cause yet more instability, and so on and so on in a vicious downward spiral; it took almost half an hour of my practically shouting at him (NO!!! DON'T BACK OFF!!! AAAGGGHH!) before he finally stepped on the gas at corner entry (by that point, we were crawling so slowly there was no way we could crash). The effect was astounding -- even my completely non-motorhead roomate (great guy, but a much better chemist than a driver) broke out into a wide grin as the previously "dangerous" car hunkered down and shot through the curve; after that he fed in lots of power on every tight turn, and we made great progress through the Berkshires...)

Where was I? Oh, yes -- great point, blacka5! When we combine your friction circle effects with how the circles themselves grown and shrink with forward/back weight transfer and camber changes (you can think of tire camber as making the circles long & skinny, or better yet of becoming laterally asymmetric), we'll have a much more complete picture of how a car's handling balance changes.

 

[nicklockard]

How do we make a MkIV better?

Replace it with a MkV...