How To Brake?

[beez]

Originally Posted by FTS

Good points Christian. Although I understand your points, IMO "smoothness" is not being smooth with your inputs but with the resultant behaviour of the car. Most pro drivers we see driving seem to have very jerky steering and other inputs, but as long as their actions result in smooth weight transfer through out the car, that is all that matters.

I disagree - smoothness of inputs directly translates into smoothness of the car on the track. Often what you're seeing with in-car camera on pro drivers is them being constantly on the knife-edge - they are moving the wheel to make fast corrections that are necessary to keep the car on the road. Emulating this style while not having to make corrections because of understeer or oversteer conditions just wears out your arms and tires.

[Larry Herman]

I too fall more into the "ramp up the brake pressure" rather than the "hit it hard" technique, but for a few as yet unmentioned reasons. First, if you've ever driven an non-ABS car, you will appreciate the additional rear brake bias that you can run when you don't initially pitch the car on it's nose. Secondly, how "fast" you can actually ramp up the pressure depends on how stiff your car is. With a softer suspension like a stock GT3 has, you have to allow time (a few milli-seconds) for the front to dive under initial application. This will prevent excessive pitching and actually allow for greater initial braking. Even with full race suspensions, I have found it better to squeeze on the brakes rather than just pound on them. The more evenly you can keep pressure on all 4 tires the more grip that you will have, and the harder you can brake.

[FTS]

Originally Posted by beez

I disagree - smoothness of inputs directly translates into smoothness of the car on the track. Often what you're seeing with in-car camera on pro drivers is them being constantly on the knife-edge - they are moving the wheel to make fast corrections that are necessary to keep the car on the road. Emulating this style while not having to make corrections because of understeer or oversteer conditions just wears out your arms and tires.

Let me put it in another way. I define smoothness as getting the weight transfer to any part of the car as fast as possible just at the right time it is needed while staying on the optimum tire slip angles. To achieve this smoothness of inputs are irrelevant, actually abrupt inputs sometimes (or may be even often) are welcome. I am of course only stating opinion here and my understanding on how the theory should be translated into practice.

For example, we see pro drivers with very abrupt steering and throttle inputs, not to mention how they use the brakes. Nothing I see in vids can be classified as "smooth" driving from the input perspective.

So, if you are in a corner with increasing lateral Gs, which increases lateral weight transfer and vertical and lateral loads on tires, the slip angles of the tires will change and that change is constant. To manage to stay at the optimum slip angles that generate the most grip, you have to modulate steering and throttle constantly as well. That is why we see such rapid and abrupt steering inputs.

Similarly, the harder you accelerate, the more abrupt the throttle, when you lift off, the more weight will transfer to the front at a faster speed. If you do execute this at the right moment, the less brake you'll need to use. But again the input is not smooth, but the resultant action creates higher grip when needed and smoother output.

Of course this is all in my very humble interpretation of car dynamics and driving theories.

[Festina Lente]

^
Sprint racing you can beat the crap out of the car,

^
Endurance racing you have to be smooth, and preserve the car.

You'd better believe the latter are very smooth with their "inputs", otherwise the race is over early.

I think you need a session or two with Hurley versus some of the curb hoppers.

[lightweight]

This is a great discussion. I've always been taught that smoother is faster. An illuminating place to learn about braking and suspension response is on a motorcycle.

[beez]

Originally Posted by FTS

Let me put it in another way. I define smoothness as getting the weight transfer to any part of the car as fast as possible just at the right time it is needed while staying on the optimum tire slip angles. To achieve this smoothness of inputs are irrelevant, actually abrupt inputs sometimes (or may be even often) are welcome. I am of course only stating opinion here and my understanding on how the theory should be translated into practice.

For example, we see pro drivers with very abrupt steering and throttle inputs, not to mention how they use the brakes. Nothing I see in vids can be classified as "smooth" driving from the input perspective.

So, if you are in a corner with increasing lateral Gs, which increases lateral weight transfer and vertical and lateral loads on tires, the slip angles of the tires will change and that change is constant. To manage to stay at the optimum slip angles that generate the most grip, you have to modulate steering and throttle constantly as well. That is why we see such rapid and abrupt steering inputs.

Similarly, the harder you accelerate, the more abrupt the throttle, when you lift off, the more weight will transfer to the front at a faster speed. If you do execute this at the right moment, the less brake you'll need to use. But again the input is not smooth, but the resultant action creates higher grip when needed and smoother output.

Of course this is all in my very humble interpretation of car dynamics and driving theories.

I guess I define smoothness as not unnecessarily upsetting the balance of the car. If you were to view the car driven by the pro drivers you mention seeing using very abrupt steering and throttle inputs from the outside, I guarantee you'd see a car that's twitching all over the road, not only side-to-side, and with weight transferring front to back as well. The main reason for not using abrupt steering and throttle inputs (both onto the throttle and coming off of the throttle) is you want to keep the weight transfers even, and as much in one direction as possible. For instance, when you come off the throttle very abruptly as you describe, you not only transfer the weight to the nose, but you can also have the car's front suspension "bounce" back in the other direction, garnering another transfer of weight at precisely the wrong time. The exception to this is when there's a dip or sudden elevation change where you might want to hammer the brakes as the suspension bottoms out to keep the suspension from bouncing on rebound, so the car stays settled. This will depend on how stiff-ly sprung the car is of course, and the rebound settings on the dampers... but in most cars, even race cars, there will some rebound from very abrupt release of the throttle. Side-to-side transfer of weight is the same thing - you only want to transfer weight once, if you can.

Some driving schools (I once heard that Jackie Stewart was a big proponent of this) use a car with a punch bowl bolted to the hood that contains a tennis ball on a string - the objective is to go as fast as you can while keeping the ball in the bowl.

I will say this - many times when you see in-car views of pros racing, they're doing these kinds of inputs because they're not on the line... when I'm doing W2W racing, and making passes, I'm spending a huge amount of time off the line in order to make headway through a lot of slower cars, or just trying to find a way around someone. Driving off the line, and in the klag, might necessitate these kind of inputs, but for fast qualifying laps or time trialing, a smooth transfer of weight, and keeping those transfers to a minimum is much preferable, and ultimately faster in my experience.

[NickW]

Smoother IS faster. What's also faster is keeping the wheel straight for as long as possible as opposed to having some steering input. And having little or to no toe either front or rear (keeps the chassis from binding, putting heat in wheel bearings creating hot spots).

I race laydown karts and road race sprint karts on big tracks, you had better believe steering input slows you down. Chassis setup is paramount for control. We watch everything from scrub radius to alignment almost every practice session- we can gain 1/10sec from more power, but literally a second or two if we can get closer to ideal chassis setup.

Of course all this is for naught if you can't drive... Smooth inputs at all times, smoothly to threshold braking, smoothly on throttle to WOT, smooth steering inputs to get into and out of corners quickly. Just because a car/kart can be darty does not mean that's how it's meant to be driven.

[FTS]

No doubt smooth is faster, but I am putting an argument as to whether the input or output need to be smoother, or are they the same?

[NickW]

Obviously if you stay within the traction circle you won't lose any traction, but to stay within that circle requires you to not exceed the chassis capabilities- those capabilities determine the size of the circle.

Ultimately we are concerned with getting through turns quicker, right? Instead of thinking of turns as brake-throttle input-turn in-track out-more throttle input, concern yourself with how to get to maximum speed as early as possible by track out. That usually means forgetting you have a ton of power, carrying far more speed into and thus out of the corner on a very stable chassis, and using far less brake than you thought you needed. I noticed my car is very neutral if the chassis is balanced, so if I don't give it any input via throttle, brake or steering the car is very settled and predictable. However, giving it throttle will lift the front, reducing the traction circle for the that end, and lifting the throttle will increase it. Braking will cause a more abrupt shift in weight transfer and really reduce the rear traction circle. So, if you are smooth on ALL your inputs you will always be able to maximize the size of the traction circle on the end of the car that needs it most. Now what you do with that traction, and whether you can use it to the limit, is another matter.

Of course, there are instances like in AX that you violate that concept to induce oversteer, but even AX requires a good pointy front end.

I'm sure I'm oversimplifying a lot of my concepts but the gist is there.

[csmarx]

Originally Posted by FTS

No doubt smooth is faster, but I am putting an argument as to whether the input or output need to be smoother, or are they the same?

I think this is a truly excellent question!

The point of my previous post, which I didn't manage to state very clearly, is as follows: the car being a dynamic system of the spring/mass/damper kind, needs sinusoidal inputs in order to generate linear - i.e. smooth - outputs.

Sinusoidal meaning slowly initially, then accelerating to fairly quick, then decelerating input to the steering, the throttle, and in case of this thread, application to the brake.

I like this theory because it encompasses most of the comments from from different people, and it seems to work in real life too!

-Christian

[csmarx]

Originally Posted by NickW

Obviously if you stay within the traction circle you won't lose any traction, but to stay within that circle requires you to not exceed the chassis capabilities- those capabilities determine the size of the circle.

Ultimately we are concerned with getting through turns quicker, right? Instead of thinking of turns as brake-throttle input-turn in-track out-more throttle input, concern yourself with how to get to maximum speed as early as possible by track out. That usually means forgetting you have a ton of power, carrying far more speed into and thus out of the corner on a very stable chassis, and using far less brake than you thought you needed. I noticed my car is very neutral if the chassis is balanced, so if I don't give it any input via throttle, brake or steering the car is very settled and predictable. However, giving it throttle will lift the front, reducing the traction circle for the that end, and lifting the throttle will increase it. Braking will cause a more abrupt shift in weight transfer and really reduce the rear traction circle. So, if you are smooth on ALL your inputs you will always be able to maximize the size of the traction circle on the end of the car that needs it most. Now what you do with that traction, and whether you can use it to the limit, is another matter.

Of course, there are instances like in AX that you violate that concept to induce oversteer, but even AX requires a good pointy front end.

I'm sure I'm oversimplifying a lot of my concepts but the gist is there.

You bring up a very important point which I often discuss with track newbies. I usually hear them excitedly exclaim that they were driving on the limit. I point out that "the limit" isn't a fixed quantity. A pro driver will drive the exact same car in a way that has a much higher limit. The point of learning how to drive isn't to get to the limit, but rather to first maximize the limit and then push to car to that limit. Which is what you are saying about expanding the traction circle by way of your driving.

The first time I went in a car with a professional racing driver, what was most striking wasn't how "exciting" it was. On the contrary it was how "calm" it all was - despite going incredibly fast. That was the first time I really got an insight into what smooth was, and how important it is to be even more relaxed the faster you go!

-Christian

[NickW]

Originally Posted by csmarx

Here is a theory:

I've recently perused a PDF about the physics of racing, and while it gets a little too involved in the equations for my (MIT educated) brain, it does have some interesting conclusions and corollaries. For example in the chapter on why smoothness is important, we learn that given the car and it's suspension is really a spring/mass/damper system it likes sinusoidal inputs.

Hmm, I've always known that smoothness is important, but I never really thought that smooth doesn't necessarily mean linear! So I've been experimenting with doing sinusoidal-ish steering inputs, i.e. start of slowly and then accelerate the input to quite quick and slowing it down again as I reach full lock. Its quite amazing. If you have a heavier car with a soft suspension try this. With my wife's Prius if I do simple linear input going somewhat swiftly into a turn, it leans out of the turn and almost feels like it's going to topple over - despite being smooth about it. If I try the sinusoidal input instead, it's completely different! It kind of squats down, almost leans into the turn and is far more stable. Wow!

I have yet to try this on track, but I image that the GT3's relatively stiff suspension and stiffer low profile tires will make the effect less noticeable. It'll be interesting to see.

Now back to our regularly scheduled programming.

I'm wondering if braking isn't the same thing. It would certainly explain variations on the theme that one should start by a softer application of the brakes to get the car to settle before squeezing harder. Similarly the idea of getting off the brakes in a more gentle fashion after a hard application. In fact if you combine these two you get: first soft on, then hard, then soft off. It's basically sinusoidal!

Something to ponder.

-Christian

For what it's worth that is the exact way I have to brake at T1 at Road America. The faster the speed, the more important you don't upset the car, especially if it's bucking like a bronco at 170mph!

[FTS]

Originally Posted by csmarx

Sinusoidal meaning slowly initially, then accelerating to fairly quick, then decelerating input to the steering, the throttle, and in case of this thread, application to the brake.

-Christian

Now that I have re-read your original post on this and this post, things are making more sense to me, at least this above statement is the first link, conceptually, between my chassis dynamics and driver training. Previously, I have not been able to link the two together in concept or execution.

My driving experience, as with most that replied to this thread, certainly collaborates this statement. My data also collaborates this statement. The part that is not collaborating or making sense to me is why would some professional drivers that I talked with advice me to be abrupt, especially under braking. The same thing happened during my recent PSDS visit too, which the coaches were in the camp of what Mr. Neil Roberts states.

There is no questions, I put out posts that are argumentative to generate discussion, but at the end I am trying to find out what is the next level of driving I need to be shooting for. I think the answer right now for me is the sinusodial approach (Christian, you should trademark the term "sinusodial driving" ); keep collecting data and analyze it until I can hit certain lap times while being absolutely smooth with inputs and observing the smooth outputs

Thank you.

[bman]

Originally Posted by csmarx

it was how "calm" it all was - despite going incredibly fast.

Originally Posted by NickW

carrying far more speed into and thus out of the corner on a very stable chassis, and using far less brake than you thought you needed. I noticed my car is very neutral if the chassis is balanced, so if I don't give it any input via throttle, brake or steering the car is very settled and predictable.

I'm sure I'm oversimplifying a lot of my concepts but the gist is there.

I agree strongly with this! Carrying more speed through and OUT of a corner has much to do with the stability and balance of the chassis, in my experience. Easing of the brake a little earlier than I am comfortable doing, staying CALM, can actually result in a more stable entry resulting in earlier throttle application.

But......how you use the brake to achieve balance seems to vary depending on the corner. I'd love to hear opinions on when you want a flat chassis vs a loaded nose vs early throttle?

For some reason, I developed a bad habit of maintaining some brake pressure on ALL turn in's; a habit of trying to keep the nose down for good grip. But I'm fighting that habit and realiziing I've got to try different things.

T10 at VIR and T12 at Mid Ohio come to mind as examples of where I want straight line braking and off the brake early enough to insure a flat chassis at turn in. Both seem to have some postive camber at first and then flatten near apex.......so does flat apex = flat chassis??? If so, why does T1 at VIR seem to require a lot of trail brake? It has some positive camber on the entry but goes flat near the apex.... I assume it's because you're approaching at high speed, trying to maintain speed at long as possible with a relatively low speed exit? Yes? Or does it have more to do with the radius?

I understand a lot has to do with the car but are there any rules of thumb or opinions on proper chassis attitude relative to track surface (elevation/camber/radius)?

I like this thread, thanks!

[NickW]

Originally Posted by bman

But......how you use the brake to achieve balance seems to vary depending on the corner. I'd love to hear opinions on when you want a flat chassis vs a loaded nose vs early throttle?

That's a function of how large you need the traction circle to be at whichever end you need it at most.

For some corners like at Road America's Carousel, a long sweeping slightly downhill righthander that literally lasts 10 seconds or more, a neutral throttle will help keep the car in shape, a slight lift will tuck the nose back in (gain front traction from moving polar point forward plus camber gain from front end suspension compression that results from caster amount, PLUS a larger tire contact point at the front left wheel), and a slight acceleration will push the front a bit (scrub off lateral force if you've turned in a bit too much).

Again, I'm simplifying it quite a bit, there's a lot more going on- for example, LSD locking/unlocking, the rate at which it locks/unlocks, etc. But for the most part, I think I'm not too far off in concept.

Originally Posted by bman

For some reason, I developed a bad habit of maintaining some brake pressure on ALL turn in's; a habit of trying to keep the nose down for good grip. But I'm fighting that habit and realiziing I've got to try different things.

That's the worst thing you could probably do. Don't forget about the traction circle- the more traction you use for braking, the less you'll have for turn in. You may be expanding the circle by getting on the brakes, but the effective turning traction is probably severely limited. You have to ask yourself- which is more important- braking or turning in, and divvy up the circle accordingly.

Brakes on a racecar (or street car on the track) are not for actually stopping- you never want to actually stop. Brakes are for modulating speed- a tool to get the car to behave the way you want it to BEFORE you get to the apex. I've noticed that most people tend to overbrake because they thing threshold braking into corners is the fast way- probably read it in books, but never really properly taught by someone. The function of threshold braking is far more difficult to execute CONSISTENTLY- you may be able to do it 20% of the time, maybe even 70% of the time, but to drive effectively on the track you have to learn to execute each action as close to 100% as possible. Most people I know who threshold brake to the apex (like what they teach in the Skippy books) are incredibly bad at it, tending to overly slow down their cars- you can hear their mistakes when they release the brake and get on the throttle very abruptly. Even the pros make mistakes doing this- every time you see a racecar go straight and miss the apex, the driver is probably fighting braking forces (getting the car slow enough to get within his turning traction circle).

Originally Posted by bman

T10 at VIR and T12 at Mid Ohio come to mind as examples of where I want straight line braking and off the brake early enough to insure a flat chassis at turn in. Both seem to have some postive camber at first and then flatten near apex.......so does flat apex = flat chassis??? If so, why does T1 at VIR seem to require a lot of trail brake? It has some positive camber on the entry but goes flat near the apex.... I assume it's because you're approaching at high speed, trying to maintain speed at long as possible with a relatively low speed exit? Yes? Or does it have more to do with the radius?

I understand a lot has to do with the car but are there any rules of thumb or opinions on proper chassis attitude relative to track surface (elevation/camber/radius)?

I like this thread, thanks!

I can't comment on T10 at VIR, it's been many, many years since I raced either course there and never the full course. At Mid Ohio T12 in a kart is a throttle breather- maintenance throttle before turn in and back to WOT before T13. In a car I would say that corner is a bit tricky because of the subtle off cambering of the turn (track falls away from direction of turn). The other thing to note at T12 is you have to give up some of the track out for the turn in for T13, which again you give up some of the track out to get a good run onto T14 and the main straight. If you construct your corner exits taking each turn in reverse you will usually end up with a more ideal line (at least that works for me).