Notes on Track Design

[Purist]

The "Two Straights Raceway" thread, along with some of the race recordings I've watched recently, got me thinking about some particular points of track design.

Compromised Set-Up and Slipstreaming Circuits
There has been talk off-and-on about having more circuits that are NOT so middle-of-the-road, with a little bit of everything, and requiring a moderate to high downforce load for best performance. A key component of this that has been mentioned is having more tracks that force a compromised set-up, because enough of the lap is composed of long enough straights that it is beneficial to take a significant portion of the downforce off, which then makes the corners trickier, and braking zones longer.

Of course, the likely result of this is a circuit on which slipstreaming is a primary, if not the dominant, characteristic of the racing at a track. The difficult bit is that corners often take up such a large portion of the lap time that it takes quite a bit of the lap being not just straights, but long straights, to reach the "critical mass" we're looking for in order to get that compromised set-up tobe viable, and even preferred.

Here are some circuits, and rough percentages of their laps composed of long straights (here defined, generally speaking, as a straight of 800m or more). I'll try to give a spectrum here, but also show the range of tracks that all fit within the category as well.

Slipstreaming Circuits:
Paul Ricard- 3.4km/5.8km=58.6%
Monza- 4.2km/5.8km=72.4%
Spa- 4.4km/7.0km=62.9%
Le Mans- 9.4km/13.6km=69.1%
Hockenheim (old)- 4.8km/6.8km=70.6%
Enna Pergusa- 2.8km/5.0km=56.0%
Fuji (old)- 3.0km/4.4km=68.2%
Surfers Paradise- 3.2km/4.5km=71.1%
Kyalami (old)- 2.9km/4.1km=70.7%
Daytona (24 Hour)- 3.86km/5.73km=67.4%

Average=66.7%

More General Spectrum:
Elkhart Lake- 3.5km/6.5km=53.8%
Interlagos- 2.3km/4.3km=53.5%
Catalunya- 2.0km/4.7km=42.6%
Sepang- 2.55km/5.55km=45.9%
Indy (F1)- 1.92km/4.20km=45.7%
Monaco- 1.2km/3.3km=36.4%
Melbourne- 2.8km/5.3km=52.8%
Suzuka= 3.1km/5.8km=53.4%
Long Beach (1999)- 1.68km/2.94km=57.1%
Estoril (1994-98)- 1.90km/4.36km=43.6%

Average=48.5%

At certain tracks (like Monaco), the 800m thing isn't really going to apply, and you have to be relative about it. Also, there will be the occasional straight that will count, even though by itself, it's too short, because the entry speed onto it from the preceding corner makes it work as though it were a longer straight (like a few places at Spa). And then you have "straights" or straights joined by exceedingly fast and/or flat-out corners (like the back side of Paul Ricard).

On the matter of what makes slipstreaming circuits, the critical value seems to be around 60% of the circuit being formed by "long" straights. This appears to be the tipping point for deciding to go with significantly less downforce/drag in order to maximize top-end.

And, of course, there is not a perfectly clean delineation. The Indy F1 road course is so dominated by that one flat-out section that it could be argued that that is something of a slipstreaming circuit. Interlagos could be considered a borderline slipstreaming circuit with its very long start/finish stretch, as well as the Reta Oposta immediately following. On the other hand, Spa has so many quick (not flat-out though) corners that it's character is not necessarily so dominated by the two primary flat-out runs as is the case with some of those other slipstreaming circuits.

BTW, these are not perfect calculations. There are estimations involved, but the above figures should communicate the general idea.

I will have a few more topic areas to bring up as we go along, so stay tuned.

[dyewat808]

Fantastic idea for a thread.

It would be great if we could come up with some sort of "Theory to Track Design", and this is a great starting point.

How about looking at seeing if what we consider a "good" track to race on can be quantified? Would it be possible to come up with some sort of analysis that allows us to say a certain race-track is quantitatively the best in the world. You could have certain "indicators" such as what Purist has suggested, as well as things like corners per mile, and average radius of corner (or something similar).

Corners per track could be a more easy one to start off with.

There are quite a few obvious obstacles in the way of course, but it might be nice to see what people can come up with.

I look forward to seeing what your other topics are Purist!

[Purist]

It certainly would be difficult to quantify what makes a "great" race track, simply due to the number of variables at work, and the fact that not all tracks will even have entirely the same set of variables, or parameters, to work with.

Part of this is intended as an eye-opening exercise, as there are aspects of track design that I have seen some on 10ths just utterly discard, and I firmly disagree with some of those posters' stances.

In addition though, I think there are some aspects or characteristics that can be quantified to an extent. As the opening post shows, I hope, the make-up of a circuit definitely can tell us important things about what sort of racing we are most apt to see at that course. Being able to spot those things helps us all understand our own and each others' tracks that much better.

[the_calculator]

Great idea indeed!

Like Purist and dyewat suggested first we need to establish the parameters to build up a "rule of thumb".
Perhaps we then can turn this into a challenge

Altough (as you pointed out yourself) I don't believe a straight/corner compromise racetrack is necessary to create a slipstreaming track (e.g. Dijon Prenois).
I agree it is good formula for succes.


Feel free to add, delete, adjust/refine

Parameters, value (my opinion);
1. percentage of straight (+800m) on the track, >60% (as I said above)
2. minimum length of a straight, +900m (The most important one I think)
3. corners per track, ? (I believe of no importance)
4. corners per mile, ? (imo not really that relevant)
5. corner radius/speed before entering the slipstreaming straight, ? (compare the two slipstreaming temples of speed Monza and Reims)
6. average corner radius, ? (tough one, high speed corners demand more downforce then a hairpin/chicane thus affecting you're corner/straight compromise)

Given the amount of parameters maybe it's best to leave out borderline tracks.

[the_calculator]

Sorry, Purist, I seemed to have missed you're reply.

Originally Posted by Purist

as there are aspects of track design that I have seen some on 10ths just utterly discard

I agree.

[Purist]

Now then, here is another area that I think needs some attention.

Variable-Radius Corners and Corner Sequences
This topic is perhaps not as straightforward as the first, so I will do my best to elaborate. I have noticed that many newer circuits seem to shy away from having complex sequences of corners all rather close together, like the infield at Interlagos, or just about the entire Sportscar Circuit at Sears Point (Infineon). Current designers also seem to shy away from variable-radius corners. The most notable exception to this observation is Turn 8 at Istanbul. However, even that corner isn't in the same vein as many older, variable-radius corners. What I mean by this is that Turn 8 at Istanbul has much more clearly defined apexes than you would have seen in earlier, variable-radius corners.

These changes to and omissions from current track design have major consequences. Long sequences of corners that all flow together put a premium on precision and rhythm, and magnify any errors a competitor makes, because those errors continue to cost them down the track that much further. The lack of variable-radius corners, and the type of variable-radius corners used anymore, where they are present, serve to constrict and more tightly control the racing line, which, when you're actually trying to race someone, is actually harmful, because you make any "alternative" line someone might try to use in overtaking substantially more disadvantageous.

It might seem like I show a tendency to dislike technical circuits, but the real issue behind it is what I'm describing here. Newer technical sections are usually uninteresting and ineffective.

Hungaroring is very technical for a good portion of the lap, but it does this through a number of constant-radius, clearly defined from one another, corners. Also, most of the corners are slow to medium speed even for F1 cars. It creates a follow-the-leader setup, and this is emphasized by several of these corners alternating between going left or right, which usually puts you on the wrong foot if you exit one corner and try to dive up the inside on the next. The straights are so short, and your line is so compromised, that it doesn't make much sense to attempt a serious move.

For a comparison, I have four tracks that come to mind pretty much right off that are good technical circuit overall, or at least have good technical elements, and these are a mix of new and old circuits: Jarama, Portimao, Interlagos, and Navarra.

At Interlagos, the section from the entry of Ferrardura all the way around to Juncao gives amazing scope for different lines, turn-in and braking points, and places to catch the other guys off and to get caught out yourself. The "ideal line" really depends on what you are trying to do, as well as what corner(s) you put most emphasis on. There is just enough room between the tighter corners to get a bit of a run going. It is not so much though, that the accordion effect allows the guy in front to pull too big a gap just based on the what the speed differential does to the distance between vehicles with the same time interval between them.

Portimao uses interesting combinations of corners and variable-radius coners, spaced out more than they are at the Interlagos infield, to create some of the same effects. Again, the "ideal line" is a conditional thing. Also, the higher speeds between the tight corners, coupled with the dramatic elevation changes, can put the drivers more off-balance, allowing those with better grip/handling to make opportunistic moves in the interesting places.

I can't quite put my finger on what it is about Jarama that grabs me. Part of it is probably those banked hairpins, and then there is just the distribution of quick, challenging corners around the lap that force rapid transitions from slower to faster sections as you traverse the circuit. I think I will be watching the Superleague Formula and FIA GT3 races from Jarama again here. Also, Iwould say that the esse bends at Jarama seem less obtrusive to the track flow and layout than is the case at some other technical circuits: Hungaroring or Fundidora Park for example.

I have some issues with the Navarra Circuit for other reasons, but I like the tricky, decisive moves that can be pulled off at Turn 2-3, 5-6, and 9-10 due to how these corner combinations work together. There is also just enough room to get some interesting lines coming out of Turn 13 and going through Turns 14 and 15 at the end of the lap.

One particular thing I wanted to demonstrate here, and that, if you know the tracks, you should see it, that decreasing-radius corners do NOT have to be bad for racing and overtaking. This is one thing that has bothered me with some of the posters in the F1 section here, and clearly, what they try to contend does not have to be the case at all. A number of the corners in the Interlagos infield are decreasing radius, as are two or three of the critical corners/complexes on the back side of Portimao.

It's also kind of funny, and kind of sad, that Tilke's best sequence of consecutive corners is probably one of his least liked by fans: Turns 17-25 at Valencia. It flows well, is quite fast, and provides as good an overtaking opportunity at its end as any other on the entire course.

Anyway, I think that's quite a bit to chew on for anybody, so, once again, feel free to discuss this or the first topic area, and I'll be back with another installment in the next few days here.

[Purist]

Remember, it is not just radius that determines a corner's severity, but also how many degrees it turns through, and banking or adverse camber will impact this as well.

The number of corners on circuits can give a general meter, but is going to be most useful for circuits at either extreme (i.e. those with very few or a huge number of turns). Numbers of turns per mile can be rather useful, especially when you know the tracks you're looking at, as it's unlikely you will have a track that has that many flat-out turns all in a row, or that many flat-out corners in total really.

For a shorter, club circuit, like Knockhill, having a straight or "straight" of 600-700m should be fine. For something longer, say, starting at 3.0km in length, you probably want at least a longer straight (750-900m or so). Once you're up to or beyond ~4.8km (3.0mi), you want to look at having multiple long straights in all likelihood.

[Purist]

It's been a bit longer than I intended, but here's the next piece.

Bends, Pinch Corners, and Other Niggly Bits
This area can be one of the trickiest to deal with. There are fewer constants to work with, and whether a curve works is very dependent on the track context.

Now, I probably have some of you wondering about that second item: "pinch corners". So, I'll hit that first. Basically, a pinch corner is a bend or kink located such that it pinches the effective racing line. When badly placed, these corners can have a very negative impact on racing potential. Pinch corners are also bad for bunched fields on starts and restarts, because they're a pile-up waiting to happen. I like good racing, and I'm sure you guys do too, but I do NOT want half the field wiped out right from the get-go.

Geometric corners tend to be worse for this. Circuito de Navarra is particularly troublesome with respect to this sort of corner. The big offender at Navarra is Turn 1, although Turn 2 doesn't exactly help things either. This corner really exemplifies what can go wrong. It's at the end of an 800m straight, so you're set to make a pass, or in the process of making one. This is a fast corner, but it's of a sufficient angle that you really need the full track width on exit. This is exacerbated by the next turn being even more severe, which makes getting a good, wide entry there that much more critical.And since Turn 3 is the tightest one of the sequence, not keeping the proper line from the start just begins to stack up against you with how much more you will have to slow down eventually, and how wide you are likely to run when you reach the hairpin. The final difficulty in this case is that these three corners are all very distinct turns, and separated by straight straights. If you had "straights" between them that had a slight bend, the effective turn angle is reduced, and since you're turning a little already, the vehicle is less likely to be unbalanced by the added input, which makes taking THE ideal line that much less essential.

Now, not all pinch corners are a real problem. Turn 4 at Navarra doesn't have enough of a straight leading into it for it to cause any notable problems, and you're not far enough along that run to be ducking out to pass anyway.

The worst variety of pinch corners are tight, or confined, chicanes. With these, there is just straight-lining between the first few apexes. You use basically one of two lines, and these are predetermined by the line you approach them on coming down the preceding straight. The trouble is, if two competitors are truly on the limit, they cannot negotiate those first two apexes side-by-side. One of them MUST give, or else there WILL BE a collision. Particularly bad examples of this include chicanes at Monza, Surfers Paradise, Portland, and Lime Rock (the new configuration). I'm also rather concerned what's going to happen in this year's British GP at Silverstone when everyone tries to pile into the Abbey Esse right after the start; there isn't much straight before that corner, and even at reduced speed from the standing start, I doubt more than two abreast is going to work through there.

Just to clarify something, the first two chicanes at Monza are tight; the opening chicane at Adelaide is not tight, but it is definitely confined.

I will add here that pinch corners can be useful for setting up passes by setting the guy behind up right behind the leader, on the same racing line, heading down a good straight.

Now, I'll take a look at bends, curves, kinks, and all manner of extra niggly bits that can cause difficulties. And the big issue created here is these added little changes in direction can REALLY foul up overtaking attempts. The especially happens when saide bend or curve is too close to the major turn that follows.

This issue of extra little jinks is what really plagues the urban circuit in Valencia. Turn 1 isn't a problem really, but just before Turn 2, there is an extra little bend to the right. Any overtakes into Turn 8 are made much harder by the pinch corner at Turn 6, and then you have the extra curve back to the right for Turn 7. The run to Turn 12 is complicated by right/left/right that comes just before the braking zone. The run into Turn 17 has a left/right to deal with just before that braking zone.

Except for Turn 7, none of these 6 extra little bends are really apparent on the track map. They are very subtle, but their impact is NOT. All of them serve to effectively tighten the corners that follow them. If you're trying to overtaking on the inside, you're already on a tighter line than is ideal. The extra bends just compound this, especially when you're not only closer to the apex of the main corner, but closer to the apex of these extra turns as well.

The ultra-compromised line for making an overtake is the biggest issue in all four aforementioned zones, and is the factor of note on the approach to Turn 2. For Turn 8, the trouble is made worse because drivers go from the far left inTurn 6 to the far right in Turn 7. Basically, you have to outbrake the guy as you're finishing the turn-out from Turn 7 and trying to make Turn 8, rather than just blowing it completely. The 3 bends heading into Turn 12 present the issue of the extra compromised line. On top of this though, the ideal line in effect weaves between the walls, when in fact it is almost straight, but this leads to the problem of the leading car being in the middle of the road for the regular line until you're virtually on top of the braking zone. The same thing happens with the 2 curves on the approach to Turn 17. Without the extra bends to interfere, you'd have good length of run right along the left edge of the track leading into Turns 2, 8, 12, and 17, with which you could quite reasonably set up the guy in front of you.

When Tilke added the Mercedes Arena to the Nurburgring GP Circuit, a kink was added to the front straight before Turn 1 proper. You essentially have to be coming alongside the competitor you want to overtake by the time you reach that kink and the braking zones, or else you are almost certain to blow the corner if you try the banzia it and straight-line the kink. You'll just be on too shallow and line, and that's all there is to it.

One thing all the above examples have is that the last bend or kink before the main corner is in the same direction as the main corner, but even in situations where this is not the case, the can still be some issues. Having a bend in the opposite direction of the main corner just before that corner can make blocking into the corner extraordinarily easy; just watch the FIA GT3 races from Jarama, and what happens going into Turn 1.

I know there are some other examples out there, but I think this gives a good basis for further investigation and discussion. And I'm sorry if I'm long-winded to the point of being a bit overwhelming with these. I just like to be accurate and thorough with my analyses.

[dyewat808]

Very interesting stuff Purist and an excellent read.

I definitely agree with pretty much everything you've written, especially the problem with pinch corners.

I think Tilke in particular is guilty of this (who else ) - it appears to be that he uses a pen, a ruler for the straights, and a compass for the corners.

Seeing more sophisticated corners (decreasing or increasing radius) would be nice to see in his designs, rather than the almost digital-like characteristic of most of his tracks.

In my view, the best tracks out there are those with far smoother corners, and where it starts becoming difficult to differentiate between where one corner end and another starts - that's a good circuit IMO.

With Sketchup this is definitely a problem (great a tool though it is) - if you just use the curve tool once for each corner, you get lots of constant radius corners.

[LuiggiSpeed]

Originally Posted by dyewat808

With Sketchup this is definitely a problem (great a tool though it is) - if you just use the curve tool once for each corner, you get lots of constant radius corners.

There's an easy way to do complicated, variable radius or increasing/decreasing radius with SKTH, I believe I explained it in the SKTH thread but what you do is you outline the type of corner with just straight lines and then do tangential curves connecting all of them:

[the_calculator]

OK before everybody here starts to erase their "pinch corners".
Like you said Purist;

Originally Posted by Purist

whether a curve works is very dependent on the track context.

Wich gets a bit overwhelmed by you're following story.

I think you're analyses is totaly true and well laid out, except for the conclusion.
There are lot's of examples where the "pinch corner" is actually a star on the track.
As you pointed out there is really only one racing line through these type of corners, wich make them extremely difficult to get right.
Getting them wrong means you seriously hurt you're laptimes (of course also depending on the straight right after).
One of my favorite pinch is the Acque Minerali at Imola; a full commitment corner.
Any fault there is also magnified because of the following climb uphill.
And what about turn 6 at Albert Park?!

The examples of Valencia and the Nurburgring are indeed dreadful.
But I've watched the GT3-Jarama videos you've mentioned; what's not to like about this, this is TOP racing.

I know you're stance about chicane's and you'll probably know mine so I won't go on about that.

Great reading indeed, thanks!

[Yannick]

Purist, thank you very much for your in-depth analysis. It is a very good read. And most importantly, your effort provides us on this forum with some vocabulary for things we want to say but up to now did not really know to say how. Given the fact that we are living in all directions from which the wind can blow, that is important.

Regarding the issue of pinch corners, I feel it is important to have tracks on a series calendar that do have a differring amount of pinch corners when analysed in comparison to each other.

The ratio of number of corners per track length and the ratio of straights or full throttle sections per track length are important figures to have in mind when designing a track, as well as the ratio of number of overtaking opportunities per lap time. It is here that shorter (but undoubtedly classic) tracks such as Brands Hatch and Dijon Prenois have their strengths.

I'm very much looking forward to your next entry in this series of theoretical background writing. Thanks again.

By the way, how would you define a Mickey Mouse Course? For me, it is a track with a high ratio of lap time per track length and a high number of pinch corners in comically short succession.

[ECW Dan Selby]

My 'Mickey Mouse' course would certainly be Indianapolis GP track. Horrid, horrid.

And thank you, Purist. Fascinating stuff. I agree that you've put alot of text down and explained it in such a way that many of us are probably nodding our heads reading it, saying to ourselves "Yep - That's certainly a theory I share".

Selby

[Purist]

I went back to the Jarama races I have. I think more what I was remembering was Gilles Villeneuve straight-lining the left-hand curve so he would be, by default, on the inside for the first hairpin there.

Acque Minerali is certainly a challenging corner, but overtaking there isn't all that common. Also, even if you somewhat blow the exit, there's that kink back to the left almost immediately, and then you don't have a huge run to do anything along that stretch because of the Variente Alta. If there is a 'star" corner at Imola, I'd probably says it's Piratella, as it can get pretty dicey forcing it two-wide through there and down the hill to Acque Minerali.

It's a fairly similar deal with Turn 6 at Melbourne. Turn 7 immediately after, and Turn 8, make defending on the run to Turn 9 relatively easy. And that stretch isn't terribly long, so it's not a terribly common point to pull off a move anyway. Turn 3, 4, and 5 are really what set up the unpredictability that allows for passes into Turn 6.

Probably the best pinch corner and chicane I've seen is the left/right just before start/finish on the Shanghai street circuit Superleague Formula used in 2010. You don't really slow for it, and it's soon enough after the preceding corner to not muck up overtaking attempts. It also then puts you dead astern of your competitor (in his slipstream) for the main part of the front stretch, which gives a very good opportunity for overtaking going into what is effectively Turn 1.

[Purist]

It's been way too longsince I added to this. And things could use some livening up around here anyway.

High-Speed, Technical Circuits
To some, this probably looks like an oxymoron. Considering the modern definition of technical circuits, this is hardly surprising.Internationally, the Hungaroring is one of the best-known and most referenced technical circuits. Here in the States, several of Alan Wilson's designs are lumped into a similar category. They are relatively slow, with an abundance of slow and medium-speed corners, have few high-speed corners, and often lack particularly long straights.

Some of these things will change with the generation of cars you are talking about, but these should give you some ideas on technical road courses that are quite quick. In their heyday, the old Spa-Francorchamps circuit, as well as the course at Rouen-les-Essarts, were very fast, but also exceptionally challenging.

Spa had several very high-speed curves that were real corners back when F1 ran the old layout. Eau Rouge was certainly not flat-out. Haut de la Cote needed to be negotiated properly to get a good run down the hill to Burneville, where you want to just barely lift, and then mash the gas and hold just the right line to where you never had to adjust the wheel to smoothly traverse the succession of apexes. After this, you had the bend at the Malmedy slip road. The Masta Esse could just be taken flat in the late 1960s, but took total commitment, and a lack of imagination about what would happen if you got it wrong. Around the rest of the lap, you had the banked Stavelot curve that bypassed the original hairpin in town, the right-hander at La Carriere, the then not-so-flat-out Blanchimont, and the then terrifying Clubhouse Corner (where the Bus Stop was added after the course was shortened for 1980).

At Rouen, you were faced right off the top with that series of fearsomely fast esse bends, including the infamous "Six Freres" corner. If you have seen the photos, this is the sequence where the shots were taken showing Fangio hanging the tail out at well over 100-mph in the 1957 French Grand Prix. After the cobblestone hairpin, you had a few potentially tricky bends leading into Sanson. After this, you had to get Beauval and L'Etoile just right to make good time around the entrire back side of the course.

With today's cars, the Nurburgring Nordschleife is not a slow circuit anymore. Yet, it still contains 150-175 corners (depending on how you count it), and the great majority are not flat-out. Apart from the Dottinger Hohe, none of the"straights" are really that straight, and that long straight at the end does have a sting in its tail with the gut-check at the Tiergarten. The series of corners after Flugplatz and before Adenau Forst demand extreme precision at speeds that can be approaching 180-200mph in the fastest cars, and the same applies for the run from Bergwerk up to just before the Karussel. A handful of the high-speed corners, like Pflangtzgarten, require more than just taking the "ideal line", but taking a line that isn't too compromised, but also allows you to land all four wheels back on the road (depending on your vehicle) before you must make the next turn; that particular drop-off is where Stefan Bellof flipped his factory Porsche 956 in the 1983 (and final) 1000km on the Nordschleife.

What I want to convey here is that a circuit need not be slow to be very technically challenging. I have nothing against technicality, but I am against putting in some stadium section or what-have-you just for the sake of having one on a track. If the track works as well or better without it, I'd just as soon leave it out. I've said it before, and I will say it again, not every track needs to have a bit of everything. Also, there are few things that can be as boring, unimaginative, and counter-productive to actual racing as a tight, technical infield section inserted into a track's layout. Hockenheim's original "Stadium Section"was good, the vast majority that have come along since then have been not nearly as good; many come across as afterthoughts, or, as I say, plunking in a technical sector just so you can say that the track has one. It is only their for its own sake. Hockenheim didn't need a second stadium, and the Nurburgring and Austin didn't need stadium sections at all. I rather liked the Castrol Esse on the Nurburgring GP Circuit, and Austin already has a series of esses after Turn 2, without needing for there to be a niggly infield bit after the back straight.

As a final example, look at Bathurst from the run into Quarry around to Forest Elbow. Again, this is an exceptionally challenging piece of road, and much of it is negotiated near or above 100-mph, even with the big, bulky V8 Supercars. Imagine an LMP1 lapping Mount Panorama in anger, or look up the F1 demo run on YouTube.

If I had to name a few newer circuits that are fairly technical, but also quick, these would probably be my top five: Potrero de los Funes, Algarve, Autopolis, Mugello, and Brno.

[olivk]

Purist i find your corner and general track analyses very interesting, and i agree with most of it - eventhough i'm no specialist.

One thing has not yet been much discussed and is often left out - as far as i can see - from track design in this forum : the elevation changes. For me the 3rd dimension of a track is as important as the 2 "flat" dimensions. I have only a small and totally amateur track experience - and absolutely no racing experience except rental karting - but the 3 tracks i've driven on so far made me believe a track MUST be analysed in 3 dimensions. These tracks are folowing ones :

> Mas du Clos (France) : A lovely small track with an impressive climb at the end of the pit straight leading into a 180° right turn that goes back down at the exit. You just can't even guess the feeling driving this corner when you see the track from above.

> Charade (France) : Lots of elevation changes all the way arround the track.

> Nordschleife, everyone knows it. As it is a fast track, the elevation changes come even more into account. Many of its corners get all of their interest from the elevation changes. Some examples :
- The sharp left down in Fuchsröhre can be passed at very high speed thanks to the enormous downforce you get down there.
- The brake before Wehrseifen is quite frightening as you plunge down into the valley, and tests your braking skills and your guts. I think 99% of the people brake way too early there (me included).
- The Brünnchen exit right turn, going up again, looks easy and trustworthy, but it's a tricky corner as the topography tends to give you negative downforce. The fact most crashes on youtube are in this corner is not only because most cameraguys are posted there.
- The depression in the straight just before the Pflanzgarten right turn gives you a lot of downforce and makes a hot on the brakes at this exact point very effective. If you use this well you will always have to brake behing the Porsche that just passed you over and that braked too early - even if you have only half the horsepower.
- etc etc.

It's easier to design flat or almost flat tracks, and it's easier to make them safe. But tracks with dramatic elevation changes are just way hotter IMO

This could be a challenge idea : Take an existing track layout, add or change the 3rd dimension to make something totally new.

I find modern tracks - especially the Tilke's - often quite boring because they are just flat. Yaaay Buddh has a cool T3... but hey it's only so cool because that kind of turns have become the exception. Maybe for safety reasons it's not possible anymore but i really wish we could see new tracks that could rival with Spa. I'm sure it's not a matter of designer's talent.

To see more "3D-tracks" maybe we have to replace Hermann Tilke with John Cameron

[Purist]

Before I get into the rest of the post, I don't think elevation change is overlooked or left out in this section of the forum. Some of us give descriptions of elevation in our track designs to varying degrees. The main thing is, putting it visually into our designs is very difficult and time-consuming. Making those hills and such look more presnetable or realistic adds a further layer of complexity. We do it from time to time, but as far as doing it as a hobby goes, we can't be worrying about all those details all of the time, or it just becomes too much. It ceases to be so much fun, when you overdo it.

Alright, since you prompted it, here are some of my thoughts on the matter.

Elevation Changes
I don't disagree with your take on Tilke, but I'm not convinced that a number of his layouts would be anything spectacular, even with some more pronounced elevation changes. I still think you need a layout that just looks "right" in plan form in order to create a truly great circuit. Now, the right arrangement of hills and valleys is what can turn an alright or good layout into a masterpiece.

The other problem we see with Tilke's efforts is the simple fact that he has a given piece of ground to work with, which may be flat as a pancake, or near to it. With his track outside Istanbul, he had a very rolling terrain to work with, and came up with a rather nice result.

It would be exceedingly difficult to create a wonderful, artificail landscape from scratch, and anyway, the real money in today's F1 circuits goes into the buildings and amenities, not into the track itself. Abu Dhabi is the epitome of this.

I actually had heard of Mas du Clos a while back, when I stumbled onto some thread here, so I am somewhat familiar with that track and its wonderful undulations.

I very much like the old Clermont-Ferrand as well. It was not on the same massive scale as some other public road circuits, but it was still an excellent course. I'm not as much a fan of the truncated layout used there nowadays.

The Nurburgring Nordschleife, well, it needs no further explanation to those who know it. Those who don't just have to see it to believe it.

Getting your elevation change "right" is a very tricky thing. For instance, you may be looking at whether to run more or less straight down a hill, or at doing a series of diving sweepers instead. You might take a more diagonal route up a hill, or attack it head-on with how you position a particular stretch.

Of course, with fantasy tracks, you can make these decisions at will, and so long as your ups and owns can be made to add up, you're fine. I certainly didn't skimp on my written description of the very hilly path carved by my design for Montello Park. I also gave a rolling ground track to one of Yannick's designs a while back.

Probably the biggest thing I can say to try and help others when they look at their designs is that variation is the key. Mix your combinations of up and down. If you can't do it visually, in the written record, differentiate between some steep and some gradual or moderate rises and falls. Having the terrain lend banking or a touch of adverse camber to your turns is always a nice touch.

It's not really about being complex either. Just try to make the elevation fit with the circuit. And you can turn a seemingly simple hillside into varied segments that provide plenty of challenge for an entire lap. Mosport Park basically goes down a hill in a couple stages, and then comes back up that hill.

If you have this gift, it can be an invaluable thing to be able to mentally construct your track in all its sinuous splendour. Being able to "see" the circuit in your mind's eye is the most powerful tool you can have at your disposal. Even if you're on the construction site of a design coming to fruition, you likely can't see the whole track if there are significant hills incorporated into the layout. So, visualizing the scene for yourself on the spot is still extremely useful.

I really can't provide any hard and fast guidelines in this area, because there are so many things that can work when done right. There are just as many ways to get a poor result also, but so much is conditional and contextual that you can't make effective, solid rules with regards to elevation change in your track designs. As long as you don't have the track going over a sheer cliff, or straight into one, your own imagination is the main limiting factor.

[ScotsBrutesFan]

Just to pick up on and add to your comments Purist...

I'd love to add elevation to everyone fo my designs, but several factors preclude this

Firstly the software that I use - Google Sketchup, allows you to create curved surfaces ... trouble is it doesn't allow you to draw on them, so you can't have the track running over the surface.

[Purist]

On that front, SpeedingTortoise has provided some guidance on how to have your tracks take in hills. It's tricky to carry out, certainly moreso than if you could just draw the track in on the hillside, but the result is nice enough once you get the hang of it.

[Himlar]

I have found one way of doing it, which I don't think has been covered here:

I do my design two-layered; on one layer I create the elevation, and on the other I design my track on a flat plate. After that, I extrude my flat track down towards my elevation layer, making sure that they intersect. After that, I use the "Intersect faces with model" option to actually "apply" the track to my elevation layer. Then it's just a lot of work left deleting surfaces and lines, and there's no way of correcting any mistakes (not that I have found, at least).

This method is complex, yes, and time consuming, yes, but if you Really want to make a track elevated sometimes, it does give you a nice result.

Speaking of track design, I should finish up and post some myself ...

And to stay somewhat on topic: great work Purist! I've read this with great interest, and will pick up on some ideas for my future designs.

[Purist]

If you could do a run through that process on a small scale, save a few screenshots at key steps, and post them with a description to the Sketch-Up Tutorial thread, I'm sure it would be very much appreciated by those who frequent this section.

I'll figure out here soon what the next installment will be on. I'm mulling over a few ideas in my head right now.

[Himlar]

Originally Posted by Purist

If you could do a run through that process on a small scale, save a few screenshots at key steps, and post them with a description to the Sketch-Up Tutorial thread, I'm sure it would be very much appreciated by those who frequent this section.

I'll figure out here soon what the next installment will be on. I'm mulling over a few ideas in my head right now.

I'll try to do that in the near future.

As a suggestion; could the topic of "sectioning" be something to write about? Are there any benefits in dividing the track into fast and slow sections, or should one try to keep the sector average speed close to the one of the whole lap?

[ScotsBrutesFan]

Originally Posted by Himlar

I have found one way of doing it, which I don't think has been covered here:

I do my design two-layered; on one layer I create the elevation, and on the other I design my track on a flat plate. After that, I extrude my flat track down towards my elevation layer, making sure that they intersect. After that, I use the "Intersect faces with model" option to actually "apply" the track to my elevation layer. Then it's just a lot of work left deleting surfaces and lines, and there's no way of correcting any mistakes (not that I have found, at least).

I'll have a go at that on one of my older simpler designs to see how it works

Sent from my HTC Wildfire S A510e using Tapatalk

[Purist]

Here goes with another segment.

"Mickey Mouse" Tracks and Elements
This was mentioned in this thread's first life, and I decided I would entertain the notion with some of my thoughts on the matter.

First off, my biggest focus is on circuit car racing, and I am particularly fond of road racing. This colors my perceptions, naturally, and while I try not to make too many broad, sweeping generalizations, go-kart tracks and tightly packed club circuit facilities often strike me as quite "mickey mouse". I have a similar feeling toward parking lot autocross activities as well, I will admit. Then again, based on the bulk of circuits posted here, I think a lot of us like the big tracks.

Of the various, major circuits in the real world, there are few that I would consider wholly defined by that "mickey mouse" tag. Two tracks that I would probably put in this category would be the 2002 Miami street circuit and the 1991 New Orleans street circuit. Most often, however, it is a particular place or section on a track that just seems pathetically tight or puny compared to the rest of the track that gets saddled with this label. The double-hairpin on the Indy road course is often cited for this criticism.

This brings me to the point that rovals in particular are regularly singled out for being "mickey mouse" in their character/layout. Of course, being stuck within the confines of a speedway is going to have its limitations, so this can be hard to avoid. It can help to have a larger oval within which to work. It is the rare roval that goes outside its speedway, and rarer still to find one of this type still in operation, such as Texas World Speedway.

Street circuits are also brought in for tongue-lashings involving that Walt Disney trademark, but in many cases, I have to disagree with its use on street circuits. For one, you can't reasonably use the same standards for street circuits as you use for permanent road courses, or everything becomes "mickey mouse". Even Monaco isn't really in that category for me. It's not unusually short for a street circuit. There were plenty of other tight, urban street circuits laid out at that time also; Pau is probably the best-known, and is still with us. And, perhaps, apart from the tightened chicane after the tunnel, none of the added corners really seem odd or out-of-place, though I personally wish Ste. Devote, for one, wasn't made so tight back in the '70s.

I think that there maybe is another undertone with the "mickey mouse" moniker, and that added tinge is a certain monotony. A number of street circuits, especially those with an abundance of 90-degree corners surrounded by uniform, grey concrete walls, have had this feature. Parking lot circuits suffer particularly from this image, and most infamous among them is probably Caesar's Palace, which hosted F1 in 1981-82 and CART in 1983-84.

Overall, I think the term is overused. There are times when I think it is warranted, but overall, I think people use it kind of as a fall-back when they don't have a clear idea of what they really mean.

If I was to give it a definition of my own, it would be something like this. "Mickey Mouse" refers to a circuit that in significant part, or in total, is unusually/unnecessarily tight, narrow, and slow. This is normally accompanied by the track being used by a series which emphasizes these qualities, because for that series, these qualities cause the racing to suffer substantially; a track can be "mickey mouse" for some categories, but not others. However, a circuit that would otherwise have this descriptor applied can have it mitigated, at least in part, by having extraordinary history and/or character.

Finally, as another example of a part of a circuit that is "mickey mouse", take a look at the 1992-95 New Orleans street circuit, and specifically, look at the smaller loop that goes around City Hall. It kind of blows the mind that they had 800hp GTPs going through those streets.

[olivk]

New proposal for Swindon Street Circuit (UK) is quite Mickey-Mouse http://www.gmap-pedometer.com/?r=5328052