Notes on Track Design

[ECW Dan Selby]

Very funny, olivk!

And Purist, brilliant stuff again mate. Enjoyed that and agree with your sentiments.

Selby

--

PS -I forgot to mention, that Texas track is incredible!! That's going to be downloaded on to my rFactor tonight, me thinks..

[Purist]

I don't have time to do it just right at the moment, but I'm going to do a new description here shortly.

[Purist]

Alright, here goes.

A Crucial Misconception
I want to go after a key argument that is often used to justify how few high-speed corners we may see on more "modern" road courses. What I mean is, car aerodynamics is frequently blamed for issues of not being able to follow closely or having a harder time making an overtaking maneuver. While I do not dismiss, out of hand, that this factor can play a part, the extent to which this is used is utterly overblown.

In super speedway racing, especially with NASCAR, where they're trying to make the cars edgier, while also messing about with the conventional draft, as well as the side draft, those big, heavy, draggy machines are quite vulnerable to the aero forces, but then, they're big, heavy, draggy objects putting off a HUGE wake, especially when they're doing 170-205-mph at the start of a run, and still doing 155-190-mph with worn tires.

I'm not going to get into that kettle of fish in this post. As I said, I want to look at this in road racing.

Too much of the time, we misinterpret a very basic phenomenon as something totally different than what it is. The occurrence in question is the accordion/concertina/rubber-band effect. Very simply, if the time gap between the vehicles in question remains roughly constant, the physical gap varies with the speed at which they are traveling. On the other hand, if you maintained a more or less constant physical gap throughout the speed range, it would mean that you have lost time in the slow areas, and narrowed the gap in the high-speed sections.

Now, here is where this can become a useful illustration in practice in a circuit's design/layout. Take Mexico City, and you're going to compare the impact of using the "stadium section" and the unmitigated Peralta(da) corner. What's going to happen?

So, for an Indy Car, it reaches 150-mph on the Recta del Ovalo. Making liberal use of track limits, it drops to 75-mph entering the stadium. The car pushes back up to about 100-mph, drops to 30-mph for Foro Sol, picks back up to 90-mph, and finally falls to 45-mph to make the hard right out onto the middle of the Peraltada. And for this exercise, maximum speed on the Recta Principale reaches 180-mph. So, we go from a spread of several car lengths to the trailing car having its nose just a fraction of a length off the tail of the leader in Foro Sol, and for the turn out onto Peraltada (45-mph) there is right on one length between the leader's backside and the pursuer's nose. Now, I should note here, with the transponders in the nose of each car, the time gap at that 45-mph is, in fact, representing a lead of TWO car lengths, not one; in this situation, we cannot treat the cars simply as point particles. This is where the distinction becomes quite critical, because with a set time gap, that physical gap grows from two lengths to EIGHT lengths at 180-mph heading for Turn 1 (seven lengths from tail to nose). That's what has to be made up.

Going back, using the full Peraltada, the IndyCar reaches 160-mph on the Recta del Ovalo, and drops to 120-mph at the apex, before it accelerates to that hp/drag-limited 180-mph at the end of the straight. The trailing car has a gap of three lengths from its nose to the leader's rear wing, and this is reduced to two lengths, but in time terms, a gap of three lengths, around corner apex. At terminal velocity at the far end of the Recta Principale, that gap is back out to 4.5 lengths (3.5 lengths of the visible, physical gap).

It should be clear, then, that making up twice the gap to just nudge the leader's rear end, and almost twice the time gap, is going to be MUCH more difficult. Also, entering the straight at a higher speed and in closer proximity allows the slipstream to take affect sooner, and therefore magnifies its impact several times over the case in which you've just come out of the stadium.

Getting back to the slow stuff again, there is another issue to examine. If you go from 180-mph down to 45-mph for a corner, and at terminal velocity, you had one length between you and the guy in front (two-length time gap), but were unable to make a pass, in order to NOT run over the guy, you are FORCED to give up time to maintain a physical gap between your cars. Thus, if you have repeated starts and stops, and then a long straight after yet another slow corner, it's no wonder you end up with a MASSIVE physical and time gap between cars out on that straight.

It's no surprise to me then that you NEED these 0.65-0.8-mile-long straights at F1 circuits when the setups for the straights are like that. You've set yourself up for it from the start. I might add, it REALLY does NOT help your land efficiency when you HAVE TO stretch out your circuits like that.

I had another look back at some old F1 turbo-era races at places like Zandvoort, Osterreichring, Kyalami, Silverstone, Brands Hatch, etc., and competitive cars certainly could stay close enough through the high-speed corners to keep things interesting. Keep in mind, too, those F1 cars were wider than the present ones, had some pretty honkin' substantial rear wings (i.e. wake turbulence). and were running on narrower tracks than what we have today.

If you want a more precise comparison, watch the Mexico City races for the A1GP series from the 2006/07 series, and compare that with the 2007/08 races when the series used the ChampCar chicane before entering the Peraltada. The difference it makes in how the field spreads out, and the passing that can be done, is quite apparent.

Hope you guys find this informative and useful. Also, I won't blame you if you're tired after reading this; I'm tired after having written it.

[SKG]

That was quite the read! Unfortunately I missed this whole thread as it was happening so it might take some time to catch up!

The points you've raised here are really interesting, I've never quite thought of following and passing cars this way before. I shall think on this more...

[S griffin]

If anything a wider car helped overtaking, as it meant a bigger hole in the air for the car behind

[Purist]

It may help for slipstreaming on the straights, but then again, your car is bigger as well, so you need a larger hole being punched in the air compared to today just to have the same relative affect.

And for the other stuff, like the wake in cornering or the narrower track, the wider car probably doesn't help.

Back then, at least, blocking wasn't as bad.

[AoB Special Stage]

Originally Posted by Purist

It may help for slipstreaming on the straights, but then again, your car is bigger as well, so you need a larger hole being punched in the air compared to today just to have the same relative affect.

And for the other stuff, like the wake in cornering or the narrower track, the wider car probably doesn't help.

+1

[Purist]

SKG, you don't have to save it up all for one go. If you have thoughts on some of the earlier posts, feel free to put up those responses.

A Further Consequence

Another conceptual note on my previous post would be this thought. If you have circuits set up such that you frequently have those situations where, unless you make the pass rather quickly, you're continually giving up a fair amount of time, it seems like you're actually encouraging the drivers to NOT push in the race. That strikes me as an attitude the many would find antithetical to the whole idea of even being a racing driver.

While I'm here, I thought I'd provide a few more examples.

The Bad

At Shanghai, Turns 11 and 12, and to a lesser extent, the Turn 14/15 hairpin as well as Turn 16, create the overtaking problems I discussed previously. You have these multiple, slow corners that spread the cars out once they hit the subsequent straight. Now, the back straight is long enough that you may well be able to pull back up and make something happen, but the front straight often doesn't quite have the distance you need to set up to make a move.

At Sochi, the two main stretches are at least long enough that you have a decent chance of making something happen. However, the front stretch of 1,300m plus, with its braking zone at Turn 2, is preceded by no less than four slow corners in succession (Turns 15/16 and 17/18). Also, Turn 13 is slow, and the runs from T13-15 and T16-17 are relatively short, so you won't make up much time if you've been caught behind someone and had to give up time because of it.

The gaps suffer through Turn 3 as well, due to how slow Turn 2 is. The back stretch would also be helped if Turn 10 were slightly quicker, but it's not as bad as coming off of T18 and having to slipstream well past the pits. I might add, if Turn 13 had a less acute apex, the effective added racing room there would help overtaking in that area.

The Good

The first corner at Istanbul is quick enough that the cars don't lose touch too much, and so, the Turn 3/4 zone is retained as an overtaking point with some real potential. Turn 7 is is also fast enough, with sufficient room for line variability, that if you can get in behind someone, and stay fairly close through Turn 8, you have a shot of making a move into T9.

The opening corners at Singapore aren't fast, but they aren't really slow, either. Turn 5, in particular, is more open than a normal 90-degree-intersection turn, and allows for some different lines through it, which is why the run to Turn 7 that follows works so well. Furthermore, Turn 22 requires you to slow down some, but you're probably not going to lose any time there because of the compression, and it puts you just a bit closer to the guy in front for the sling through T23 and onto the pit straight, which aids overtaking there.

On the flip side, having the more open hairpin leading into the slower one at the start of the lap at Sepang is key to making that combination work. Not having that huge compression all at once, and the corners set up to provide an array of possible lines, actually helps overtaking happen there, AND also happen then on the run around the T3 sweeper into T4. The configuration of T1/2 mitigates the spread you see between the cars coming off of Turn 2.

Well, that was more than I initially planned to do, but hopefully this helps flesh out a few more things in the minds of readers.

[Yannick]

Thank you Purist for your analysis of why all of us on this board prefer tracks which have more medium and high speed corners than slow corners. In my early days of designing tracks in the early 90s (on paper), my role models have been Monza, Silverstone (when they still used just the perimeter circuit), Östereichring and the now largely deconstructed Autodromo Nelson Piquet at Rio de Janeiro, 3 of which I knew from Accolade's Grand Prix Pursuit game of circa 1988. The one thing that unites these is that they have lots of medium and high speed corners and that they have straights. Many of those corners go around 180 degrees or more, such as Peraltada from Mexico City which you have discussed in detail. The original Hermanos Rodriguez is also an amazing layout but with the highway just behind the wall of Peraltada at the apex, they really had no other choice than to remove it from the layout if they wanted to race F1 cars there. Watkins Glen also deserves to get a mention and to be put in the same group as the turbo era tracks you have listed.

It is interesting to note that there is a difference in likelihood of overtaking between the end of the long front straights of Iberian Tracks (e.g. Estoril and Barcelona Montmelo) and the "classic" circuits you have mentioned. I believe the reason for this is how the rest of the track compromises the setup of the car, as the Iberian Tracks usually have tighter and slower corners in the infield than immediately surrounding the straight, whereas the "classic" tracks do maintain roughly the same level of speed in most of the other corners on the track, often having more than one straight on which overtaking is possible.

That is also why Dijon-Prenois does not fit in with the group of Iberian Tracks: it does have the front straight being the longest straight as defining geometric element of the track like those have, but the part of track behind the pits is unique in that it does not have any slow corner at all. This is a circuit which benefits greatly from not having that bit of everything in its layout. The DTM race that they held there a few years back was so amazing to watch, yet still, the track still does not host any major series today anymore.

Another interesting aspect to discuss might be the influence of medium to high speed 180 degree corners on how racy a circuit can be. Look for example at how much overtaking Barber Motorsports Park provides with the current IndyCar chassis. Overall, I get the impression that the car setup for Barber must be somewhere between that for Mugello and Sonoma, with Mugello having the undulating terrain and fast technicality which you have discussed in detail in a previous posting on this thread, and Sonoma having the heavier braking zones of some corners.
I've never been a fan of Sonoma but the current IndyCar chassis does provide good racing at this place, especially in the circuit's current version. It's too bad that SMI seems to generally disregard the need to upgrade Sonoma's safety features.

[Purist]

I'm a bit curious why you chose Dijon-Prenois to compare with the Iberian circuits. However, it is indeed a fine circuit, without a truly slow-speed corner, and yes, that DTM race was excellent. (I have it in my video library, and I'd love to see DTM or FIA GT go back there.)

I'd say the closest French circuit to those in Spain and Portugal might be Le Mans Bugatti.

Jacarepagua is alright, but it's never been a real standout track for me, probably because it's so flat and comparatively featureless in its surroundings, especially when looking at it next to Interlagos.

Watkins Glen, at least for the long course, doesn't have quite the same simplicity of some of those other circuits I mentioned. It went out just as turbos were coming into the sport. Also, I haven't been able to find a full-length race recording of any of the F1 races there, so I don't have as good an idea of exactly how the period cars raced there.

Peraltada at Mexico City? I would have installed a SAFER or Tecpro barrier on the outside of the corner, and gone with it. They work on the ovals, or in the tight spaces of Monaco.

Estoril, Jarama, and even Jerez, are in a different vein than Catalunya and some of the others. Estoril, before the MotoGP makeover, was quite fast and flowing. Jarama isn't as quick, but it all fits together well, and has a good mix of features. Jerez, especially without the chicane, while geometric, is still rather flowing and quick. Seven of the 13 corners are taken at high speed.

Among the newer Iberian circuits, I far-and-away prefer Algarve, which, while it's no old Estoril, is far less slow and fiddly than, say, Valencia (Ricardo Tormo).

In terms of long, 180-degree (or more) corners, the first to come to mind is likely the Carousel at Road America. I don't know that you can call any of the 180s at Barber "conventional". All three (or four?) of them are kind of split up in one way or another.

I'd guess the downforce level at Barber is probably closer to what you see at Sears Point than what you see at Mugello. You want the car trimmed out more at the Italian circuit than at either of the other two. And Sears Point not only has the harsh braking zones, but several high-speed corners as well.

The DW12 chassis has done wonders for the racing almost everywhere, and Barber is very possibly the most graphic example. However, Sears Point has not seen the same benefits, and really, once the field gets single-file after the first lap or two, the racing hasn't particularly improved. The simplest thing they could do to help is to run the old version of the Turn 7 hairpin, like what IndyCar did in 2005-07. Honestly though, they just need to run the original, 2.52-mile layout. The racing would be miles better, and the pole average would be up around 130-mph, rather than 112-mph for the current iteration.

As for safety, SMI has done quite a bit of work at Sears Point. I do think that a few of the tire walls could be positioned better. Also, I think they were better off just having the front stretch run out onto the dragstrip, rather than constricting it, and the kinks of Turns 12 and 1, between solid walls, like a street circuit. Not to mention, if having a wall right on the outside of Turn 1 is acceptable, I don't see how they can complain about a lack of run-off at Turns 10 and 11.

You might want to go to YouTube and see what the track looked like for the 1997 Save Mart 300 (NASCAR Winston Cup), before much of the re-work began. The coverage of the 2000 ALMS race from Sonoma shows things at a somewhat intermediate stage, but with the full road course still essentially as it was.

(If I go back and watch the 1987 IMSA GTP races from Sears Point and Laguna Seca, I think, as for the track itself, I might have a somewhat higher regard for the Sonoma circuit over the Monterey course.)

[Purist]

Sears Point-Racing Difficulties of the IndyCar Layout
I thought I should add some more specifics about why the original layout at Sears Point would be so preferable to the current configuration used by IndyCar.

Part of this goes back to my recent post about the compression in slow-speed corners, and then the spreading of the cars on subsequent straights. The other big piece is how the shape of the inside of the corner, as they use it, impacts the usable racing line, and how viable an inside, non-traditional, overtaking line can be.

With the 2.385-mile layout they currently use, Turn 7 is 35-40-mph. The slowest point in the Turn 9 chicane is 45-50-mph. The Turn 11 hairpin is 40-45-mph. All three have a very sharp, almost angular apex setup with the alignment of the curbing. This severely pinches the racing line, and unless the guy on the outside gives it up, an attempt at an inside move will be ineffective; that inside line is just too tight. Alternatively, if the issue is forced by the overtaking driver, there is a pronounced cross-over in the two lines being taken, and a distinct chance or even likelihood of a collision.

And then, after the harsh compression, the leading driver pulls out a significant margin in terms of space coming off of Turns 7, 9, and 11. However, unlike F1 circuits, Sears Point has no really long straights, and there isn't the room available to make one that would be long enough.

After Turn 7, you have "the Esses", and so you don't have the straight to set up, pull out, and pass until the approach to Turn 9, but that isn't long enough. The run through T10 to the T11 hairpin isn't long enough either, and still probably wouldn't be even if they went all the way down to the original hairpin turn. The spread coming off of T11 then makes it difficult to make up enough ground on the run past start/finish and up the hill to have a good shot from Turn 1 to T2 of making something happen.

So, with the old Turn 7, the entry is wider in turns of being more open with a smoother curve to it. It's not as abrupt and doesn't narrow the line as much. It's also at least a 55-mph corner, if not 60-65-mph. On the original layout, T9 is a flat-out sweeper approaching 170-mph, and the chicane is a non-issue. You get a clean, fast run through Turn 9 and the gut-check Turn 10. You have the chance to pull up and set up the guy in front for the braking zone at T11, which is 50-55-mph in its unmodified form. This reduces the spread between cars coming off the corner, and the run through Turns 12 and 1 up to T2 is also lengthened a bit, which further improves the racing potential in that section.

None of the other corners are slow enough to be a particular issue, and in its unaltered form, the flow of the track is simply exquisite. If the cars can stay together, Turns 2 and possibly 4 become much more viable as overtaking zones. And as I mentioned earlier, the increase in average speed will be quite noticeable. Coupled with the improved raceability, the action can really come flying at you, perhaps even more dramatically than at Barber. You'll certainly have less time to catch your breath between segments of watching the cars being flung through the corner sequences than you will at Road America or Watkins Glen.

[Yannick]

Thank you Purist for your extensive reply including your detailed discussion of Sonoma.

First of all, yes, I agree that Le Mans Bugatti would fit in well with the Iberian Circuits. And my favourite of the circuits in that part of the world is also Algarve/Portimao.
For me, Jerez does not fit in with the group of the Iberian Circuits either, mainly because of its flowing nature and it having more than just one long straight.

Watkins Glen is a very nice layout indeed. It's just that this track has got some serious safety issues, the most blatant being the lack of runoff surrounding the Esses, which would be necessary to prevent a single-car accident turning into a large pile-up from happening there after a start or re-start. Unfortunately, ISC failed to do something about this during their fairly recent renovation effort at the track.

The SAFER barrier idea for Peraltada at Mexico City is a good one which would have worked good enough for IndyCars because they are chassis with crash structures that are solid enough for high-speed oval impacts. Yet, F1 cars are frailer than that and it would not have been safe enough for those. Also, we both know that the allure of filling the entire baseball stadium on the corner's inside with paying spectators and not having to build new stands was just too high a chance for the promoter to pass up: taking the original Peraltada out of the layout pretty much was an economic decision.
However, the 2nd stadium section of the new Hockenheim, the Mercedes Tribüne, has also been constructed due to an economic decision, and now, Hockenheim has been suffering at the box office from an overabundance of seating capacity for the F1 race for years. They even had to give up their annual calendar spot and share it with another track because of it. People now prefer the Mercedes Tribüne over the Motodrom because they can see the action at the hairpin at the Tribüne and the Motodrom does not have any action on display anymore because the low downforce layouts are gone, which the cars used to run in the old days of the long straights though the forest.

I also agree that Sonoma is very much a stop-and-go circuit these days. Yet, the most recent layout works better than what IndyCar ran on there when I returned to watching the series after reunification. However, the less sharp and more circular hairpin which can be seen in the video from 1997 would indeed be interesting to watch again. Yet, SMI does not seem to want to adress the real problem of Sonoma, probably because it would need some major investment: the final corner and its lack of runoff due to the location of the pit entrance. At this point, it seems rather unlikely that the stop-and-go nature of the 2nd half of the lap of Sonoma is going to be changed anytime soon.

Still, I have come to the maybe controversial opinion that IndyCar is currently better off with Sonoma than it would be with Laguna Seca, as Sonoma is more undulating and has several interesting sections whereas Laguna is basically just the hill with the cork screw and the first corner at the end of the front straight.

[Purist]

At Watkins Glen, I think one of the tunnels goes under the track between Turns 1 and 2. Widening the verge AND having to do a new, longer tunnel would be a serious undertaking. Also, there's the topography to consider; the ground falls away substantially outside of Turn 2 up to Turn 4. Furthermore, there definitely is a tunnel under the track between Turns 3 and 4. Really, a significant amount of that paved run-off outside of Turn 1 should be returned to gravel and grass, so that the cars are funneled back on to the track sooner, rather than drivers scrambling at the last second, and rejoining just before the barriers come in on either side.

With Mexico City, I don't exactly like it, but I can't argue with your conclusion for F1. It's funny though, because CART/ChampCar started out using the stadium section in 2002-05, but then, gave it up.

There can still be action in the Motodrom at Hockenheim, but yes, the more moderate downforce settings probably don't help; the shorter straight beforehand definitely doesn't. Also, I think the old corner entering the stadium was slightly banked, but the new one, with the dragstrip there now, is essentially flat (no camber).

I don't know that the current IndyCar layout at Sears Point works that much better than the one they used in 2005-07. You have to take into account the fact that this latest iteration was first used in 2012, so we don't have a direct comparison on this course with the old car. In two of the spots where the 2008-11 layout was tweaked for 2012 (Turns 9 and 11), the new configuration may be a bit better than the second one.

I'm not sure what all the issues might be with redoing the pit entry at Sears Point. It looks like they own the land at that end; there's a go-kart track a little further down. It may partly have to do with NASCAR's shorter lap, and they DON'T want a pit lane that is so long, time-wise, that there is a greater risk of losing a lap to the leader if you make a stop.

I might add, the Sprint Cup cars carry more kinetic energy into Turn 11 than the Indy Cars would. Also, the sportscars, especially the LMP900s/LMP1s in ALMS and the GTPs of IMSA before that, carried substantially more kinetic energy into Turn 11 than the current Indy Cars would. I don't see the DW12 approaching that hairpin corner at more than 175-mph. The sportscars weighed more, but could still do 180+mph into there. The NASCAR machines can do 135-mph into that turn, and weigh more than twice what the IndyCars do.

If I'm looking at the IndyCar layout of Sears Point, I'll take Laguna Seca. If they can get the Indy Cars on to the 2.52-mle layout at Sonoma, I'll take it over Monterey.

[Yannick]

Thank you Purist for another extensive reply.

I've had a look again at the Esses of Watkins Glen and there is one tunnel more than I first realized there was.

If you look at this aerial view http://www.aboveallohio.com/images/s...erial_view.jpg on the right, there are the Esses. There are tunnels going through under the track both before and after Turn 4. Both would have to be lengthened so runoff for Turns 2, 3 and 4 could be built on top of them.

When talking about access roads blocking the introduction of new safety features, the problem of Watkins Glen is relatively minor when compared to other seasoned tracks.

For example, the traditional Donnybrooke perimeter circuit of Brainerd International Speedway in Brainerd, Minnessota, where USAC ran one of its first road races back in the 60s has got only one entrance for spectators enabling them to enter the infield and to see the far end of the track from inside. That road leads over a bridge across the track, and that bridge stands in the place where a fast corner's runoff should be these days.
Instead of changing the access road, Brainerd chose to abandon this portion of the track completely, and built a somewhat slower speed shortcut that includes even a new pitlane, as the old pitlane area on the former front straight is now used as a drag strip. That must have been quite an investment.
If Brainerd International can do it, then why not Watkins Glen, as the Glen's problem would even be easier to fix?

[Purist]

The access points are just the start.

Like I mentioned, the ground outside the track falls away along that stretch. They have sort of cut the corner in adding some space on the inside of Turn 3. However, for the rest, you'd have to build up a massive hillside, keep it in place until the vegetation can hold it there (it's a rather rainy spot), figure out what to do with those access roads, and that's IF the track owns enough of the land on that side to do all that, not to mention, whether the NY EPA would even approve it.

Not all the ugly incidents at the Glen happen in "the Esses".

2008:
https://www.youtube.com/watch?v=6unt74Gw96s

2009:
https://www.youtube.com/watch?v=2-thoBzdwYE

So, they add more paved run-off, and guys use it for granted, which contributes to this.

2014:
https://www.youtube.com/watch?v=Jc4QX_FjeZs

(The SAFER barrier is why you guys have NO run-out at the final corner. An "antiquated track" is fine, if you drive respectfully; if you don't, not even the wide-open spaces of Riverside or Cleveland would be "enough".)

I think there are two tunnels under that stretch of the track. One leads to the interior of "the Boot"; the other leads to the outfield side of the track. The land drops away too much within "the Boot" to make messing with that tunnel anything less than a total nightmare.

As for Brainerd, the bikes had a slower Turn 9 built for them to largely mitigate the bridge issue on the old course. I think the bigger "problems" people see with the old course are the banked Turn 1 and the use of the dragstrip for the front straight. Then again, since they're easily off of Turn 10, and hit the strip in a straight line, that shouldn't be such a big concern.

(And regarding the new course at Brainerd, I just have no great desire to watch much of anything run on it.)

[Purist]

I'll move this one back up, since that other thread is now floating out there.

[Purist]

Harsh Compression/Expansion Zones
Some of the recent F1 discussions got me thinking about this. Basically, these features occur where you have a slow corner, and in the most extreme cases, it's a tight corner between two straights of some notable length. So what happens is the physical gap between cars rapidly decreases, but then shoots right back up again as they transition from the braking zone to the acceleration zone.

This phenomenon, when it occurs too frequently on a circuit, tends to have a real detrimental impact on overtaking potential, because it is this feature that leads to the rapid spreading/separation of a field of cars in just the first few laps after the start or a restart.

Now then, I've calculated as best I can the frequency values in miles per instance of these sorts of zones for all 21 current F1 circuits. So here are some figures.

Overall Average: 0.748 miles
New-Build Circuits: 0.670 miles
Non-Permanent Circuits: 0.608 miles
Modified/Mutilated Existing Circuits: 0.669 miles
Comparatively Intact Classic Circuits: 1.016 miles

And just to give a wider lens, here are some of those figures applied to not only the active F1 venues, but also F1-capable circuits that aren't on the 2018 calendar.

New-Build Circuits: 0.774 miles
Non-Permanent Circuits: 0.584 miles
Modified Existing Circuits: 0.612 miles

And here's a special bonus.

New-Build Circuits Never Used by F1: 0.802 miles

There do appear to be various features that can influence how much of a separation you see between the cars when negotiating a slower corner. Circuit flow and rhythm play a role. One of the most obvious factors though seems to be elevation change. The steep downhill run after La Source at Spa looks to negate a substantial amount of the expected growth in the physical gaps between cars. On the other hand, the largely uphill run out of Spoon Curve at Suzuka appears to somewhat frustrate attempts to close the gap on the car in front heading toward 130R.

I'm continuing to compile a listing, including a wide array of non-permanent circuits from across the globe, as well as a data set for more classic road courses.

[LPP68]

Originally Posted by Purist

Harsh Compression/Expansion Zones


There do appear to be various features that can influence how much of a separation you see between the cars when negotiating a slower corner. Circuit flow and rhythm play a role. One of the most obvious factors though seems to be elevation change. The steep downhill run after La Source at Spa looks to negate a substantial amount of the expected growth in the physical gaps between cars. On the other hand, the largely uphill run out of Spoon Curve at Suzuka appears to somewhat frustrate attempts to close the gap on the car in front heading toward 130R.

That does not seem to make sense. If the widening of the gaps is mainly caused by hard acceleration after a slow corner, than increasing that acceleration by having the acceleration zone on a downhill slope would only increase that problem, and keeping acceleration intact by doing it uphill would decrease it.

[Purist]

My understanding is that a downhill run reduces the impact of a poorer corner exit. Derek Bell talks a bit about this while narrating his lap of the Nurburgring from 1983 on "In-Car 956". Also, with that example from Spa, you can see it with your own eyes that the gap generally doesn't open up as much as you'd likely suspect from coming off a corner as slow as La Source.

Beyond that, as I talk about, it's a compression AND expansion zone, so the corner entry plays into it as well.

[LPP68]

Originally Posted by Purist

My understanding is that a downhill run reduces the impact of a poorer corner exit. Derek Bell talks a bit about this while narrating his lap of the Nurburgring from 1983 on "In-Car 956".

That would make the effect even worse. Many overtakes are only made possible when the driver in front makes a mistake under pressure. Cornerexits witch do not punish mistakes with time losses on the next stretch will decrease the chances for overtaking.

Also, with that example from Spa, you can see it with your own eyes that the gap generally doesn't open up as much as you'd likely suspect from coming off a corner as slow as La Source.

Except that you wouldn't suspect that. The whole point why Tilkes designs offer so many slow corner - long straight - slow corner sections is that modern high downforce racing cars can only follow eachother closely in slow corners, since in fast corners the second car is losing to much downforce negotiate the corner as fast as its opponent.

Beyond that, as I talk about, it's a compression AND expansion zone, so the corner entry plays into it as well.

I have no comment on this.

[Purist]

Please read my earlier posts in this thread, titled "A Crucial Misconception" and "A Further Consequence", which deal with the accordion effect, that is, the relationship between the time gap and physical gap as the cars go up and down throughout the speed range around a circuit.

No, it doesn't make it worse. Like I said, you can see it on display. Perhaps I was a bit crude in my initial explanation. The gravitational constant, is, well, a constant. When exiting a corner, if you get a better launch, you are, in theory, going faster at the same point on the track. When exiting a corner downhill, gravity is working with you, but because you have resistance from rolling friction and aerodynamic drag, you're actually getting relatively less help from gravity than if you had a poorer launch.

I'm sorry, but I do suspect that. What gets ignored with slower corners is the accordion effect, and its impact is seen because two objects cannot occupy the same space at the same time. That means, the leading car necessarily reaches the acceleration point sooner chronologically than the trailing car, and the physical gap opens between them as a result.

Downforce-producing aerodynamics get unduly blamed for the perceived difficulty in following another car, but you're missing something essential here. In a high-speed corner, the apparent accordion effect will be smaller; so the amount that it looks like the second car is closing in on the leader mid-corner is reduced. Remember, if you have a slow corner where the cars are immediately nose to tail, and you have a corner with triple the apex speed where the nose of the pursuer is two lengths behind the leader's tail, the time gap in those two scenarios is the same. That means that the cars are actually running closer together than you think when negotiating a high-speed corner.

As for the problem you get with what Tilke does, well, have you ever noticed how far apart the cars get on the first half of the first back straight at Abu Dhabi? There are multiple slow corners right before that straight. You have a situation where if you can't make the pass into the chicane, you may be forced to relinquish time in order not to hit the back of the leader at the hairpin, but then the physical gap blows out heading onto the straight, and you need a lot of space for the slipstream to tow you back up to the car you're trying to pass.

A couple of other things can/do help a trailing through high-speed corners. The first is the simple fact that generally a driver isn't going 100-110% allo-out all the time, unless you're Ayrton Senna, so you usually have margin to simply push harder than the guy in front. The second is that, no matter how much downforce the car makes, there is a finite limit to how much grip the tires can actually give. So even if you're losing downforce in the leader's wake, if you're still making enough that the tires are giving you all that they can, it doesn't matter.

[Purist]

Hope I didn't come off as mad or something. I was mainly tired, and maybe a little exasperated.

The main thing with a downhill run is that, the faster you're going at a given point, the closer you are to the terminal velocity you could reach on that slope when propelled only by gravity. So the better your corner launch, the less gravity will help you. To put it another way, when going uphill, your relationship to gravity might be somewhat akin to the act of pulling a rope, but on a downhill run, well, how much good does it do to push on a rope?

I hope that helps.

[LPP68]

No, I don't understand why you think downhill exits help to overtake. I do get the accordeon effect though. Basically the question is If the negative effect of this effect is bigger than the lack of negative effect of losing downforce in a fast corner. Tilke clearly thinks not, for F1 cars at least. For bikes and touringcars the results are much different ofcourse.
I do suspect that the likes of Vettel, Hamilton, Verstappen cum suis are all very close to 100% most of the time.

[Purist]

Gravity works as an equalizer; so with falling objects (going downhill), the tendency will be to negate the speed advantage that one of the two has. Also, when you're going downhill, you effectively have less resistance to push off against, but in turn, that means your useful effort actually goes down.

Going uphill is harder, but getting that better push off at the start will be rewarded more greatly for the whole of the subsequent flat-out run. It's the same thing in cycling; the uphills are what break up a group and can yield large (time-wise) breakaways.

And modern engines have less of a tendency to give out than a biker's legs, so once you start that pull out of, say, Juncao, it's relatively harder for anyone to pull up to you than on the flat, even with a slipstream.