F1 circuits the root of all evil? *Updated with #4*

[Ciro Pabón]

Well, what can I say? First, thanks for the thread, again. I really enjoy this. If only I wasn't forced to defend Tilke... I have my own criticisms, but they are different from the ones raised by our respected members. However...

Second, I have a few minutes now, after working my butt for the whole week, so, even if its 1 am, it's time to relax with all of you.

Let's start with this: a track design is, normally, not just an attempt to trace a line, Scot, with all due respect.

Yes, I know you probably spent a lot of time imagining what will happen when you race on your first "attempt". However, let me explain a couple of things, for you to counterargue.

You know I refuse to acknowledge "authority" by itself. Everything can be explained, so even if I work in things related to this thread, it means nothing if I cannot explain them in simple terms. That's how "engineering thoughts" work, that's the (wonderful!) spirit of this forum. So...

You normally start a track design with a lot, a certain terrain with borders. You cannot (normally) go to the owner and say: "hey, I want to buy three more acres", or "hey, I don't like your land, sell it and buy a new one".

So, you don't simply start with a blank page, but with a site, a definite place on the surface of Earth.

You have to walk the site, understand how the water moves in it and plan your drainages accordingly. The very life of the surface depends on it. Yes, you can commit a lot of "sins" with curves, but if the pavement is not durable, then the owner is in trouble (and your colleagues will make a few jokes ).

Few people understand how expensive is to build a kilometer of road (around a cool million devaluated dollars for a simple 12 meter wide road, no accesories included, specially "works of art", that is, extra-drainage), but all drivers understood that a poor pavement is a danger and not exactly the better thing for racing.

For example, this year, next week, Indianapolis is having its 100th birthday, something I will celebrate (and I think Scotracer will too ). In its first attempt Mr. Francis surfaced it with compacted gravel saturated with oil.

First pavement at what wasn't "the Brickyard", yet


What did they get? “Drivers were quickly covered in dirt, oil and tar. As the track surface disintegrated in the turns, flying gravel shattered goggles and bloodied cheeks. Driving at Indy was like flying through a meteor shower.”

Of course, nowadays we have other methods. But all of them include a judicious analysis of drainage, specially on flat terrain. Water must move, you do not want it to accumulate under the track. Why? Because it will show very quickly you're not a good designer!

The weight of racing cars is insignificant for a road, compared with the weight of trucks. Internal strains in a road are proportional to the fourth power of the load. So, racing cars that weighs 600 kilos, that is, more or less 1/100th of what a truck weighs, cause only (1/100)^4 damage. One truck passing causes as much damage as 10 million F1 cars...

So, in a racing track, any bump or crack comes from bad drainage That's how important drainage is (I won't go into expansive clays, that's another matter, nor into the spring thaw where many of you live, or into tropical rain, where I live).

Tell that to the people that build Gilles Villeneuve on an island! As you can imagine, water level is high there... thus, the problems with the surface they had. Same thing you can observe at Monaco harbour...

So, any creek, rivulet or water current you find at your site, you have to "respect". Besides, you have to understand and model how the water under the surface will move. It's not intuitive. That's what I was talking when I spoke of "subtle flavors". If you're working on a mountain, the problems are not smaller, but different.

In the same way you could design a wing or a car body just because it seems "nice" or it appears aerodynamic. Actually, you need a wind tunnel and some modeling, don't you?

The main problem is that you also have very different orders of magnitude into play, the same as with loads.

Water can take a few seconds to move through one meter of gravel, but it will take centuries to move through one meter of clay. The whole science of soil mechanics didn't advance one centimeter until Mr. Terzaghi understood that, during the 1950's...

You can predict settlements with the tools he invented, but it takes time and money: again, it's not a blank paper, you start by probing the soil. This is a large restriction to most designs.

Even if you design in a desert, well, then wind comes into play. You have to find some way to protect the surface from sand. Besides, building a road on sand is, well... the proverb about building a house on sand comes handy. Just lemme tell it's not easy. It's actually harder to do.

You also want to know where the sun is, not only for visibility of drivers, but for spectators not to cook under it.

You, first of all, are not building a track for drivers, but for spectators. I'm sure you jumped quickly over regulations about bleachers, pitlane and garages, not to mention parking lots... but they matter even more than the track.

So, trying to be short (as if I could! ), what's left after you consider a few things, that have nothing to do with racing, is the area where you can design the track. You will be surprised at how narrow it is sometimes... I swear.

I guess is the same thing with cars. Probably you're full of ideas about downforce when you start, but after you finish studying the rest of regulations, you find that you're very restricted about what you can do.

Well, it's 1:30 am. I'll continue tomorrow, not because I think somebody is interested in reality checks, but because I somehow feel obligated and because this thread could become a nice summary of what you need to do to build a good track.

Imagine how restricted you are at Valencia, please. It's people falling asleep Tilke's fault entirely? I don't know, but surely he's not to blame for having to build a circuit around a dock! Try to do it and we talk later.

Now, if you could tear down downtown Valencia, you could probably build a better track. Superimpose Scotracer track on it and voilá. Sorry, sorry, Scot, it's just a lame joke, I couldn't resist...

One more thing (after using the word "obligated", straight from spanish ): altough I can find the word "envolvent" in some english documents I have, I should have used the world "envelope". Sorry, Andartop.

I'll try to explain where I think the magic of Mr. Tilke resides by explaining how this envelope is used (by him and nobody else, except, perhaps, Wilson Motorsports and Alpha), and the trouble it implies with the location of stands.

The simple phrase in regulations about "designing the track following the racing line" (or something like that) is where the problem resides.

Final: it's not minimum radius what is the result of V2/K, Scot. I'm sleepy right now and I'm not into the mood to read rules at this hour, but I'm sure vertical curves are parabolic, they're not circles.

I'll check regulations tomorrow but I'm pretty sure that this formula is used to deduce the minimum length, not the maximum radius, unless I'm deranged (I could be, yes, I know ).

[Scotracer]

I know you begin with a piece of land but here's where Tilke does his thing: he bulldozes it to get a blank (as blank as one can be) canvas then build his creation on that. It's funny how the areas he's been given/chosen are of a very flat variety just outside a major city.

And I haven't commented on the Valencia street circuit as I know that's a street circuit so complaining of the layout would be pointless (same with complaining about Singapore).

If I had a 3D terrain modeller I would attempt to create a circuit (in fact, as a child I would take an area around my town and try to fit the best circuit on it) but I don't so that was just a rough-hand circuit layout. Of course it has to be adapted to the real world.

I know I am only scratching the black-top (;) ) with my analysis but if it has been done in the past with great circuits such as Spa, Clemont Ferrand, Nurburgring etc it can be done again - with even better construction techniques.

Turkey is probably the best Tilke track and will gain character but it is still a wide-open circuit. We need more tunnels =D>

You don't have to explain this all to me; I am an engineer but I just fail to see why the current designs must all follow this very similar pattern.

Here's a flying lap of Abu Dhabi I did on rFactor:

[youtube]http://www.youtube.com/watch?v=TB6oLxWMs9k[/youtube]

As you can see, it is just created from Tilke set-pieces. Whilst obviously in a desert you can't have hills and valleys, you can still create a unique layout!

[TheMinister]

I think perhaps that 12m track width is more important than it appears. Think about the most exciting, breathtaking corners or complexes (spa turns 1->4, wall of champions at montreal, mirabeau, turn 1 at interlagos all spring to mind for me) and all of them are on fairly narrow track. If the track is wider, the corner will be faster and have more room for error. Of course with wider track you can just make the turns tighter to maintain the same racing line, but it doesn't work in a complex.
Some other factors they have in common:

-they all punish mistakes. One little slip and your car is damaged, or everybody comes piling past you. It doesn't have to be race ending, just that an error should not preserve your track position.

-hills (aside from montreal)- changes in elevation make it more exciting, not sure why.

-overtakes not final. All these turns are such that a move down the inside at the start of the complex does not guarantee you are through, and the person being overtaken can easily come back at the next turn/on the exit.


Contrast all that tilke circuits- he designs overtaking zones where it's one simple move and you're through, and an error is completely OK. This just doesn't make for exciting racing.

[gridwalker]

Seeing as we're talking about elevation changes and gradient : the first time that I stood at the bottom of Eau Rouge, I was taken aback by how much steeper it was than it appears on the television screen. Does anybody know what the steepest gradient is on the current F1 calendar?

[jonathan189]

The Tilke tracks are by no means the worst. Sepang and Shanghai gave us great races this year, and they look fantastic with their signature grandstands. They are no worse than Barcelona or Silverstone in terms of facilitating good racing.

Also, Istanbul has one of the best corners in F1.

[Ciro Pabón]

Seeing as we're talking about elevation changes and gradient : the first time that I stood at the bottom of Eau Rouge, I was taken aback by how much steeper it was than it appears on the television screen. Does anybody know what the steepest gradient is on the current F1 calendar?

Well, I think you just saw it at the Grand Hotel hairpin, (formerly known as Loews hairpin and before that as Station hairpin), in Monaco.

I estimate it has around 20% of gradient (which is barbaric ) in the inner part of the curve, because of the transition of the sideslope, which can increase the "true" centerline gradient by 8 or 10%, tops.

If I try to measure it, using the sea horizon in the background as a horizontal line, I get around 27%, but I cannot believe my own figures. Something must be wrong with the image perspective, or it is an insane slope, pick your choice. Few roads have above 10%, most tables for calculations go up to 8%.



About the lap, Scotracer, I don't know what to think of rFactor, even if its an excellent car simulator.

Check the video: it seems to me that you lost the car twice during that lap (in the real world) by putting a tyre on top of the strip of grass.

I'd say that something is wrong with the simulation and that it does not take in account the difference in grip between asphalt and grass or sand, but, of course, I might be wrong.

This seems to me like a lost front


I also would like to know how could I build the hills and valleys at Yas Marina being 500 meters away from the sea, on a sandy soil.

Perhaps by importing the gravel (and clay to make the base cohesive) from Iraq? (just another lame joke ).

After reviewing the site for hours while writing about Yas Marina for a different thread I'm pretty sure you're at least tens (if not hundreds) of kilometers away from good quarries for landscaping. Clay does not seem abundant in the area (you need water to have it!).

I'm not saying that is impossible to create hills there, or that I know the exact location of sources for materials in the site, of course, but I can tell you that if your quarry is, I don't know, 50 km from the site, the cost of landscaping is twice the "regular" price for quarries up to 20-30 km (wich is the "norm") and that landscaping the whole track means landscaping also the stands and everything.

However, before starting the two quick posts I've made in this thread, I was trying to explain (in a long writing I've not finished) that is very important to try to find a natural amphitheater.

I'm not against changes in elevation, I'm just trying to explain that there are sites where you can use them and sites where you cannot.

My example could be Spa, but I love the theater at Epidaurus, as an example of how you try to exploit a site to create good stands.

Theater at Epidaurus. Those trees behind the stands give you an idea of the "natural" shape of the site. Ideal!


I think is almost as important as changes in elevation to create stands that allow you to watch the whole (or a majority of the) circuit.

Brands Hatch (or any american oval!) is a notable example of a track where you can watch almost the entire track from the grandstands (instead of relying on a large screen, or worse yet, watching only one curve from your 500 dollars seats).

Brands Hatch natural amphiteater (well, for England, which is not precisely the Himalayas). I would cut down the trees... but probably I would be killed by the local Greenpeace chapter.


For a nice 3D quick and dirty modeller you could try Google Sketchup. It's really easy to use. Also, any CAD package has the ability to model in 3D, but it's harder to do quick sketches.

[Conceptual]

I love that corner in the 2009 Indy car in rFactor. It is actually really fast since it lifts the inside front tyre off the ground, and allows you to three-wheel it around very quickly. I actually put in a 1:14.008 on the F1RL Monaco track with that car, and that is really something for a car that never was intended to be driven on that course!

Anyways, Ciro, is 27% really unbelievable? I mean, that street was more than likely laid down long before the engineering and science that drives modern roadways was even conceived. I am sure that there are many crazy things like that around the world... I just wouldn't expect to find one in the middle of a F1 track!

[Conceptual]

Ciro,

What would be necessary for a track to require a optimumally symetrical car setup?

[Ciro Pabón]

Well, you need to have as much curves to the right as curves to the left, if I got your idea of neutral setup. So, you need a bridge or overpass (like in Suzuka) to create a figure-8 track. Did I understand you? I'm not sure.

About the slope in Monaco, I'm not sure also. As I said, the sideslope increases the "true" longitudinal slope. I looked for the critical site, where the sideslope transition "push downwards" the inner edge of the road, so...

Monaco is an ancient track, it gives you an idea of why I complain constantly of old tracks not being "updated to the speeds" of modern cars.

Driving an F1 car in Monaco has been compared (by Coulthard?) to piloting an helicopter in a living room.

Imagine what you should do to update the Nurbürgring Nordschlieffe, with its cliffs to the sides of the road... or its vertical curves designed for 80 kph speeds...

I'm not against nostalgy, but reason must find a small site in our collective brain.

Now, I concede that if you have the money to build a new road around Nürburgring, you can build a track site there without being accused of murder.

Even tracks more modern like Catalunya offer a poor show to me. I like tracks where you can separate good drivers from bad ones, and in the current F1 environment, where drivers have been trained since childhood, you need to create subtler tricks, that's all I'm saying.

Sorry, it took too much time to comment a few things, I haven't been able to give a good explanation of why Tilke team is excels in some designs (in a sense) yet... perhaps tomorrow.

[modbaraban]

Check the video: it seems to me that you lost the car twice during that lap (in the real world) by putting a tyre on top of the strip of grass.

I'd say that something is wrong with the simulation and that it does not take in account the difference in grip between asphalt and grass or sand, but, of course, I might be wrong.

This seems to me like a lost front

That's nothing wrong with rFactor. It's just that the modders have probably failed to dial in the right grip parameters for every surface (this track is a modification). Isn't that an artificial grass, btw? I thank we see cars touching it at times.

[Ciro Pabón]

Well, yes, modbaraban, that's the point. If the simulation can have that kind of mistakes I assume there are no differences in grip for marbles nor sand.

However, Scot fantastic video is a good example for me of the kind of mistakes you can make when simulating. Sure, it is a simulator for a game, but here you have some other examples:

Are the kerbs that flat, or is the suspension (nor grip) responding to them (or is it me)?

Are there severe problems with the sideslope (or am I becoming blind)?

Do F1 cars tilt sideways like old Cadillacs or is the sideslope at curve 5 (and several other sites) totally wrong? (red line shows the horizon)


Is the track unevenly heated at any site or is it uniform? For example, nor the stands nor the palm trees project any shadow.

You can count the ways in which the track simulation is inaccurate. From my point of view, most simulators don't give a damn about tracks (or I am mistaken, who knows).

Now, my point is that when you design a track, those subtleties give it its taste and make a lot of the differences between Alonso (who could be aware of those subtleties) and Piquet (who might not be).

To judge the track only for the lack of fast curves (without a lateral G force simulator, not yet invented, for you to evaluate how dangerous they are) could be naive. That, or Tilke team is composed of morons. As a norm, I do not assume that our colleagues are stupid (and that attitude goes against the ethic code! ).

Now, can I complain of armchair drivers without offending anyone?

[modbaraban]

To be short I'll say that you are correct on most ponts aboout track sumulation.

But it's not the point. We haven't seen F1 cars racing for real at Abu Dhabi. But we have seen them in say Sepang.

Most people on this thread (viewers, fans not track architects!) would agree that there's nothing spectacular about a car taking a complex and the only optimal line through the wide and seemingly endless ocean of tarmac called Sepang.

Just compare this:
[youtube]http://www.youtube.com/watch?v=9mPHx4j2xks[/youtube]

to this:
[youtube]http://www.youtube.com/watch?v=ElSIY8AChFA[/youtube]

Obviously Toyota Yaris is a better car than a BMW E30. But which one is more exciting? Same with the tracks.

[Ciro Pabón]

Sorry for trying to counterargue again.

The math theory for trajectories, which I "ate" in its entirety, with all my gullibility, says that you can deduce an optimal trajectory for one curve (with ease, almost using a simple calculator) but that you cannot deduce the same for a set of curves... That's the theory.

So, based on that, some designers say that if you build a wider track you have more overtaking opportunities, because there could be "hidden" optimal trajectories that an inspired driver could discover.

This forum has seen me trying, naively, to "invent" a way to "consolidate" those ideas, without much success.

You can see at some Tilke designs certain "desperation" when he tries to "balance" a good set of curves, not too fast (where nobody will dare to overtake, at 4 Gs and 300 kph!) nor too tight (where your only chance is outbraking the adversary).

To my disappointment, I haven't seen just one overtake at that kind of "mixed" curves, while in days of yore that was where the action happened.

I guess that point is what has given us the designs that Scot despises...

So, after watching the results of simulations and the actual results of races, after watching Tilke experimenting with the whole issue, I'm not sure about the theory anymore, as the videos you post make more or less clear.

"Mixed" curves can be a blessing for "underfunded" genius drivers in stock cars, but maybe it's not the same when you drive F1 cars.

As usual, it's easy to blame "excessive" downforce.

If you ask me, there is no solution yet, because it is a deep question.

On one side you have "the NASCAR experience" where you throw out the window the track (1 mile wide! 30-40% sideslope!) and the car (all of them are the same!) because you hold that it's all show management. On the other, you have FOTA claiming that technical development (and market management!) is the true reason why they compete.

I think that the "show management" solution is the dangerous path Mosley (and thus Tilke) has taken with tracks and that's my objection to their strategy.

Check this: "The issue is to keep the manufacturers involved, to keep F1 racing as an elite part of racing and to make the technology relevant. A Ferrari, a Porsche, AutoUnion, Alpha, Lotus, they didn't become the names without the F1 heritage. We cannot take that away and make it a spec series like the IRL". Check this (very intelligent) comment at 5 minutes into this clarifying video.

[youtube]http://www.youtube.com/watch?v=eh4KyJUE8y8[/youtube]

Now, with tracks is happening the same: there is a spec for tracks, so narrowly defined (when you take in account the regulations and the requirements for the "accesories") that you're driving away the "old AAA clubs" from F1, clubs which used to be the base of F1. Silverstone and Jack Stewart come to my mind.

Sure, Mosley needs to "push away" guys like James Stewart or Nicholas Craw, because they truly represent the drivers and the fans, like Scotracer or you or me.

Nothing easier than squeeze them and replace them with a sheik or a chinese ruling party that doesn't know about karting or simulations or anything, because they are a new power base for the elite we have had ruling FIA for the last 20 years.

On the other hand, when you replace Tilke with some other guy (like Alpha, for example) the solutions they arrive to are similar: they use the same mathematical tools and similar theories of design.

So, I think that blaming Tilke is like blaming McLaren for keeping itself silent through the current "spec lowering" debacle.

That's why Montezemolo is fighting with all its might about the "limited budget", that's why Mosley is fighting back.

The former defends its firm, the later defends the money of FOM and its own power base. I cannot wait for the results, this is going to be a huge clash.

Not until the track owners take a similar stand (when will that happen? I cannot foresee it really happening) we will see truly original tracks. Certainly NOBODY asks for tracks that are relevant for road design! (well, I do).

I think that track regulations should be changed, and, in my opinion, they should involve local clubs, but my reasons for that are as boring, convoluted and impractical as the ones I've given here.

[timbo]

I think that track regulations should be changed, and, in my opinion, they should involve local clubs, but my reasons for that are as boring, convoluted and impractical as the ones I've given here.

But how they must be changed?
When I play a sim and try a track like Zolder - it's great, and the biggest challenge is that it is narrow and fast. But there's no way back - narrow track = no overtaking these days (maybe except Monza...).
And what about safety?

[Scotracer]

Ciro whilst I enjoy your posts I must point out that your ramble about the geology/geography of the Yas Marinas area is irrelevant because I said myself under the video "Whilst obviously in a desert you can't have hills and valleys, you can still create a unique layout!" But it still does bring on the question: why are they building one in the desert, in the first place?

Also the rFactor video is just a "first-look", it is not meant to be an accurate depiction of the circuit (I would have to wait for a laser-scanned version for that) and that's why in general the kerbs aren't as severe as you might expect and the surface is uniform.

As a final point, I would question why the FIA demand that any new circuit be made to those specific regulations, which basically denies the creation of a new Spa, Silverstone or Suzuka. Sure, they want to make something that will stand the test of time but the circuits are only really useful for F1 cars (with the exception of Malaysia) as they are so wide, long and dominated by 1km+ straights. Of course they have the varying layout circuits (Shanghai has a great number of possibilities) but that in the end isn't the same circuit. Let's hope that Donington is a return to what is true and good.