2014 pace vs. 2004 pace, where, how are they better?

[Miguel]

Probably find that with a screaming 900hp V10, those cars back then didn't worry too much about drag.

Huh? Of course they worried about drag. Drag squares with speed, at the kind of speeds that F1 cars go at, 800, 900 or 1200hp doesn't make an enormous difference to top speed.

I'm with SiLo on this one. There was a massive difference in wing size between what was run in 2005 and 2006. This was especially noticeable in the circuits with long straights, but where aero was still quite important. Some examples would be Indy, Canada & Spa. Spa is even more noticeable, since the lower power meant cars arrived Eau Rouge slower.

[Moose]

Probably find that with a screaming 900hp V10, those cars back then didn't worry too much about drag.

Huh? Of course they worried about drag. Drag squares with speed, at the kind of speeds that F1 cars go at, 800, 900 or 1200hp doesn't make an enormous difference to top speed.

I'm with SiLo on this one. There was a massive difference in wing size between what was run in 2005 and 2006. This was especially noticeable in the circuits with long straights, but where aero was still quite important. Some examples would be Indy, Canada & Spa. Spa is even more noticeable, since the lower power meant cars arrived Eau Rouge slower.

Instead of blindly being "with" someone, lets do the maths...
power: 900hp (675000W)
air density: 1.2041 (20°C air at sea level)
speed: 210mph (93.88m/s)
P_d = \frac{\rho v^3 A C_d}{2}

I get A C_d of an car with that much power able to do 210 mph is equal to 1.35503951.

Plug that into an 800hp car instead, and we get a top speed of 202mph
Plug it into a 1200hp car instead, and we get a top speed of 231mph.

So we see that dropping 100hp only cost us 8mph in top speed. Gaining an extra 300 would only get us an extra 21mph. As I said, not an enormous difference.

[mrluke]

A 900bhp turbo car has far more power for more of the time than a 900bhp NA V10.

[Reca]

It would be nice if reca could post a sound data acquisition comparison from a track wich has the same layout now and then...

Here we go, tracks that didn’t change layout, had dry qualifying both years, and, most relevant, that I could find onboard videos of 2004 laps...

The 2014 laps you can find at formula1.com in the videos section, the others I found on youtube.

Bahrain (remember 2014 was “night” qual and 2004 was track’s debut):
https://www.youtube.com/watch?v=fJUd2Nxr2U4


Hockenheim:
https://www.youtube.com/watch?v=3__4XNcGIbg


There would be Monza too:
https://www.youtube.com/watch?v=Ev6Q2yx5bdA


but the video has a few seconds of badly disturbed signal around Roggia, so I only did a more superficial speed vs time comparison, still good enough to qualitatively see the difference both in cornering speed and straight line performance.

For instance, Ogami himself already posted here informantion he had from "the horse's mouth" that, in 2006, Renault and Ferrari were achieving 5.5Gs on Turkey's T8. And this number is quite feasible to me going by Kimi's pole lap there, in 2005.

It may sound feasible to you but reality suggests otherwise, as from my analysis of data during 2006 pole lap Massa didn’t even get to 4.5g (peak 4.35 at 265km/h):


5.5g would require something like 290+ km/h, no way a V8 powered car could get there (for instance in 2010 at full throttle the RB6 did “just” 275)

Remember that while advantaged by grippier tyres and a further year of evolution of these aero rules, the 2006 cars also had the downside of 200+ hp less than in 2005.
That already increases sensitivity of laptime to drag reduction, moving down a bit the target of optimal downforce level in a given track (higher aero efficiency can allow to get back bit of that, but 200hp is big loss to be recovered via efficiency, let alone in a single year); just because df level of 2005 was theoretically attainable, it doesn’t mean it was affordable.

On top of that T8 is in a speed range where drag already starts to drain not small part of total available power, and generating the required lateral force via tyres demands quite a lot of power too, without considering that, due to reduced acceleration rate out of previous corner, approaching speed in 2006 was considerably lower than with the V10’s.

All that combined means that even if theoretically the car could generate a lateral force as required to match (or even beat) the cornering speed of a 2004-2005 car, there just wasn’t enough power to reach/sustain it.


Notice that the same reasoning applies more often than not to 2014 cars.
In fact, while it’s true that max acceleration seen is “only” 3.5g or so (little more at Suzuka’s S’s) hence not really impressive in absolute, it’s also true that that level in most of tracks is reached already at speed well under 200km/h.

The problem in that to see higher lateral accelerations you need, first, proper high speed corners (not many left in this calendar, the T8 we were talking about for example isn’t there anymore) and second, the power to do so.

For instance, taking Bahrain as comparison, the corner where the F2004 reached the highest lateral acceleration is the one at about 3750m in the graph I posted above, turn 12.

As you can see there the F2004 is considerably quicker than the W05 was; the little piece of info missing from that graph though is that Rosberg was at full throttle there, from exit of turn 11 and all the way thru T12.

Which means that the real difference there is power.

It’s the extra power of the V10 that allowed MS to gain way more speed in that short acceleration out of turn 11 and, in the corner itself, to keep accelerating (even if obviously at lower rate) while generating the required lateral force.
Rosberg with lot less power (and some extra weight) gets there lot slower, and when he demands lateral grip, the power required by tyres to do so causes speed to become basically constant, the V6T just doesn’t give enough power to generate lateral force and accelerate at same time, as the V10 would allow to do.

[hollus]

Thank you so much Reca, that is fantastic!

I had assumed that 2004 cars were much faster than 2014 cars in all corners, but your graphs nicely prove that it is not true. 2004 cars win by a fair margin in high speed corners, but in slow corners both cars are closely matched and often the highest apex speed is in favor of the 2014 car. This makes sense, as with low downforce the extra weight stops being a penalty. The tires have more weight on them in 2014 exactly to match the extra lateral force needed. And slow corners have more influence in lap times. To me that explains why 2014 cars are not that far, weight corrected. They are only disadvantaged in certain parts of the circuit.
As for acceleration in the straights, a 20014 car with DRD deployed can match the acceleration-top speed of a 2004 car (more or less, of course it is speed variable). The time gap goes flat in DRS areas. 2014 cars lose in other areas, not only in acceleration, but also in top speed, which is almost not influenced by weight. So the gains in drag haven't been that spectacular over 10 years, even after losing so many wing elements in the back. Probably this shows the smaller diffuser hurting performance.
Also, there is no magic catching up of the speed traces after 300Km/h as would be expected of the speed selective stalling that the cars are rumored to have, so I am very tempted to call BS on this theory that flow structures are optimized to stop producing downforce at very high speeds. Only in the Bulls, maybe?

Again, thaaaank you for that post!

[Tommy Cookers]

yes, thanks to Reca !

...... The G-meter in 2004 read 6 in the fastest ones. ..... .

in the later example kindly linked by Artur I can't even see the steady 4.5 g that he sees
the display is jumping about like a toad on a griddle (why wouldn't it?)
without proper data post-processing to cancel dynamic and sampling errors, and including correction for roll etc

is there footage showing more centripetal g than this example ?
and more braking g ?


@ Harsha
the HANS presumably relieved the load on the neck in cornering
the neck load component due to the reclined pilot position in the Lavi fighter apparently gave insoluble problems
(head restraint being impossible as it would prevent the pilot eg looking behind)

[Vary]

I can't remember if in 2004 they had TC, but it seems to be years in these videos. I think it is a major factor in better laptimes

[Ogami musashi]

@Reca:
Nice post
I'm not familiar with how sound data acquisition procedures. I assume there's a model in it. Do you have any observed telemetry data to fit your model? In other words, do you have any estimation of simulation error?

Your graphs show that 2004 cars cornered significantly, sometimes vastly, faster in medium and high speed corners.
It show less differences in low speed.

This is quite what i expected, since downforce is the main difference between the two generations. Low speed cornering do not suffer as much especially as mechanical progresses happened.

As for the 5.5G, i think Artur craft may have mixed two things. The infamous "from horse's mouth" IIRC was the quote of L/D ratios of F1 cars at monaco. Again IIRC my various discussion with a tyre engineer from michelin (then in F1) was that more 3.5G in lateral Gforce was very unlikely and thus the 6g's i quote are from the displayed G meter that surely overestimates G's and/or articles so of course overstating happens.


@Hollus: I'm sorry, but i still fail to see why you come to use that weight correction. Please feel free to explain me.

Also i'm not convinced by your "more weight on tyres" explanation, as intertial mass increase has the double drawback of increasing the needed lateral force while decreasing the tyre's friction coefficient. And statistically, if you agree that anything low speed is below 150km/h, in barhain we have out of 7 LS corners: 3 in favor of 2014, 2 in favor of 2004 and 2 equal, in Hockenheim out of 4 LS corner we have 3 in favor of 2004 cars and 1 equal and in Monza the two low speed corners obviously have noise on their signal in 2004. Considering, measurement/modeling approximations, the fact that barhain 2004 was the first run on the track, the conlusion is very difficult to draw and if you discard the latter, hockenheim is in favor of 2004.

[Artur Craft]

So wait, Newey says that the RB6 probably had the most downforce of any f1 car ever made, yet rather than listen to one of the worlds most renowned aerodynamicists who knew intimately how well the RB6 performed and exactly how much downforce it did actually make, you think it is more likely that he got his words muddled up as you dont believe that the current cars look as fast in the corners?

He said so and sort of demonstrated it by means of Lateral-G achieved but other cars have already achieved 6Gs so, that alone, would contradict what he claimed.

Mario Theissen already said they recorded 6Gs in 130R(around the 2004 era) and Barrichello already said they reached 6Gs in Barcelona's T3 in 97.

Based on measured G alone, ground effect cars would've generated very little downforce(even less than this year) as they only peaked at around 3Gs.

When it comes to the flexible-wing Red Bulls, I do. Newey could crank up as much downforce as he pleased, with very little regard to the deleterious effects of drag, because his front wings shed induced drag along straights as the tips drooped low enough to burst the vortices that form inside the end plates. Combined with the best EBD of its time, those cars made prodigious levels of downforce.

That's simply what CFD and wind tunnels allow for these days. (Can you imagine how much downforce those proper ground effects cars would have made had they been sculpted with modern tools? )

I'm far from convinced by only this. Even if that indeed would reduce induced drag considerably, they are still very restricted with floor's and wings' area of pressure, so it's simply not possible to "crank downforce as much as pleased". In Monaco they always want as much downforce as possible and pretty much neglect the drag that comes with it. They can't add more downforce because they are limited to the areas of the surfaces of pressure which are already ran at top AoA.

That car had an onion layered double diffuser that went down half the length of the floor...

This is simply not allowed since early 80s. The bottom/floor must be flat ahead of rear wheels center line.

[Artur Craft]

Plug that into an 800hp car instead, and we get a top speed of 202mph
Plug it into a 1200hp car instead, and we get a top speed of 231mph.

So we see that dropping 100hp only cost us 8mph in top speed. Gaining an extra 300 would only get us an extra 21mph. As I said, not an enormous difference.

DRS alone accounts for that 8mph difference.

For instance, Ogami himself already posted here informantion he had from "the horse's mouth" that, in 2006, Renault and Ferrari were achieving 5.5Gs on Turkey's T8. And this number is quite feasible to me going by Kimi's pole lap there, in 2005.

It may sound feasible to you but reality suggests otherwise, as from my analysis of data during 2006 pole lap Massa didn’t even get to 4.5g (peak 4.35 at 265km/h):
http://imgur.com/tcwjvXj

5.5g would require something like 290+ km/h, no way a V8 powered car could get there (for instance in 2010 at full throttle the RB6 did “just” 275)

Remember that while advantaged by grippier tyres and a further year of evolution of these aero rules, the 2006 cars also had the downside of 200+ hp less than in 2005.
That already increases sensitivity of laptime to drag reduction, moving down a bit the target of optimal downforce level in a given track (higher aero efficiency can allow to get back bit of that, but 200hp is big loss to be recovered via efficiency, let alone in a single year); just because df level of 2005 was theoretically attainable, it doesn’t mean it was affordable.

On top of that T8 is in a speed range where drag already starts to drain not small part of total available power, and generating the required lateral force via tyres demands quite a lot of power too, without considering that, due to reduced acceleration rate out of previous corner, approaching speed in 2006 was considerably lower than with the V10’s.

All that combined means that even if theoretically the car could generate a lateral force as required to match (or even beat) the cornering speed of a 2004-2005 car, there just wasn’t enough power to reach/sustain it.


Notice that the same reasoning applies more often than not to 2014 cars.
In fact, while it’s true that max acceleration seen is “only” 3.5g or so (little more at Suzuka’s S’s) hence not really impressive in absolute, it’s also true that that level in most of tracks is reached already at speed well under 200km/h.

The problem in that to see higher lateral accelerations you need, first, proper high speed corners (not many left in this calendar, the T8 we were talking about for example isn’t there anymore) and second, the power to do so.

For instance, taking Bahrain as comparison, the corner where the F2004 reached the highest lateral acceleration is the one at about 3750m in the graph I posted above, turn 12.

As you can see there the F2004 is considerably quicker than the W05 was; the little piece of info missing from that graph though is that Rosberg was at full throttle there, from exit of turn 11 and all the way thru T12.

Which means that the real difference there is power.

It’s the extra power of the V10 that allowed MS to gain way more speed in that short acceleration out of turn 11 and, in the corner itself, to keep accelerating (even if obviously at lower rate) while generating the required lateral force.
Rosberg with lot less power (and some extra weight) gets there lot slower, and when he demands lateral grip, the power required by tyres to do so causes speed to become basically constant, the V6T just doesn’t give enough power to generate lateral force and accelerate at same time, as the V10 would allow to do.

Firstly, thanks very much for that post!
The bold part sums up a pretty good argument that you constructed.

IIRC, the 5,5Gs that Newey referred to was in Turkey's T8 and RB6 was cornering at "only" 260kph(a far cry from 290) on pole lap, taken straight from speedmeter.

Also, how do you know the lateral acceleration of Massa's pole lap when FOM didn't display G-meter graphic? You can't extrapolate that solely from speed during a given corner as driving lines vary

When I saw G-meter on this year's Suzuka weekend, it showed only 2.5Gs of sustained lat-acel in the Esses, peaks of around 3G, but that's over the kerb. I only watched it live, so can't say the numbers precisely, though.

A good comparison through a very fast corner could be made at Copse where Alonso says he took at around 280kmh in his 2006 pole and RB6 was doing around that same speed in Vettel's pole. This year, during FP2 option tyre quali simulation(totally dry), W06 was taking it at 240kmh and some 235kmh on Rosberg's pole lap as conditions were not optimal.

I'm sort of a layman in this matter so, in rough/quick terms, how do you extrapolate speed from engine noise, please? and how precise is that method?

[Artur Craft]

in the later example kindly linked by Artur I can't even see the steady 4.5 g that he sees
the display is jumping about like a toad on a griddle (why wouldn't it?)
without proper data post-processing to cancel dynamic and sampling errors, and including correction for roll etc

is there footage showing more centripetal g than this example ?
and more braking g ?

it keeps within 4.3-4.5G only during the second apex. It's not sustained for the whole corner at all But that's they tighest part.

As for the 5.5G, i think Artur craft may have mixed two things. The infamous "from horse's mouth" IIRC was the quote of L/D ratios of F1 cars at monaco. Again IIRC my various discussion with a tyre engineer from michelin (then in F1) was that more 3.5G in lateral Gforce was very unlikely and thus the 6g's i quote are from the displayed G meter that surely overestimates G's and/or articles so of course overstating happens.

There was no G-meter graphic back in 2004, it was introduced in 2005. I would like to know more about 3.5G being some sort of threeshold of lateral force.

I suppose the FOM's graphs, which comes directly from the in-car accelerometers, represent it quite well and we only have to be carefull with random peaks when driver goes over a bump or kerb which shows a much greater value than true lateral acceleration.

For instance, Suzuka's 130R was taken at 290-295kmh in 2001/2002(before it got slightly modified), which required a 5G lateral acceleration.

I really would like to know more of what this Michelin engineer told you.

Also i'm not convinced by your "more weight on tyres" explanation, as intertial mass increase has the double drawback of increasing the needed lateral force while decreasing the tyre's friction coefficient. And statistically, if you agree that anything low speed is below 150km/h, in barhain we have out of 7 LS corners: 3 in favor of 2014, 2 in favor of 2004 and 2 equal, in Hockenheim out of 4 LS corner we have 3 in favor of 2004 cars and 1 equal and in Monza the two low speed corners obviously have noise on their signal in 2004. Considering, measurement/modeling approximations, the fact that barhain 2004 was the first run on the track, the conlusion is very difficult to draw and if you discard the latter, hockenheim is in favor of 2004.

That's the thing. Peak friction coefficient occurs at some 500-600kg(per individual tyre) for some tyres while at as little as 200kg for others. It's entirely possible that with more weight on low speed corners, the extra load will increase mu and thus cornering speed.

[Tommy Cookers]

in a level high speed bend there is some benefit at the apex from camber transition and a related slight compression
giving speed benefit which gives a DF benefit and so a further speed benefit
slope or slope change either longitudally or laterally will have further effects
and wind of course has a big effect
(and body roll causes overread of lateral g developed by the tyres, but favourable camber adds to that g, as does aero lateral force)

so the g achieved on track will typically be different to the g achievable in the conceptually neutral condition
aero work is presumably referenced to such a condition
since max g is so corner speed dependent, imo the above is a major cause of apparent discrepancies

[Ogami musashi]

There was no G-meter graphic back in 2004, it was introduced in 2005. I would like to know more about 3.5G being some sort of threeshold of lateral force.

I suppose the FOM's graphs, which comes directly from the in-car accelerometers, represent it quite well and we only have to be carefull with random peaks when driver goes over a bump or kerb which shows a much greater value than true lateral acceleration.

I've said from G meter and/or articles. For this specific one (the 6g's) it was from an article dated 2004 (that i tried to recover but couldn't) about 2004 cars being too fast, and i recall they said the cars were doing 6g's at some parts with drivers quotes.
However people tend to exaggerate numbers and we had a discussion here (i think the title was about Kart G forces) that telemetry as presented in media is not an accurate one.

I really would like to know more of what this Michelin engineer told you.

Not that much, we discussed about the potential of a racing tyre, he basically told me that 2.5 mu coefficient would be possible but that then the reactions of the car would be sluggish. It was during one of those exchanges that he told me those 5g's were overestimated and that anything higher than 3,5G (sustained) was unlikely. Years later, again IIRC, i mailed a pirelli F1 engineer which told me about the same thing.

That's the thing. Peak friction coefficient occurs at some 500-600kg(per individual tyre) for some tyres while at as little as 200kg for others. It's entirely possible that with more weight on low speed corners, the extra load will increase mu and thus cornering speed.

Could you please provide me some articles on that? i didn't know.

[Juzh]

It was during one of those exchanges that he told me those 5g's were overestimated and that anything higher than 3,5G (sustained) was unlikely. Years later, again IIRC, i mailed a pirelli F1 engineer which told me about the same thing.

RB6 last corner, 4+ pure lateral sustained troughout with 1 hand on the f duct.

[Blanchimont]

I'm sort of a layman in this matter so, in rough/quick terms, how do you extrapolate speed from engine noise, please? and how precise is that method?

Maybe Reca will correct me or give a better explanation, but here are my thoughts on this.

1. Analyse the (main) frequencies that the engine creates. These frequencies should be proportional to the car speed, if the tyre slip is constant. You now have graph of frequency(time).

2. A upshift will result in a drop of the frequency. If you identify an upshift, simply scale the new lower frequency to match the frequency just before the upshift.

3. Scale the complete graph by comparing the frequency at the speed trap with the recorded speed( http://www.formula1.com/results/season/ ... _trap.html ) or scale the frequency so that the integral of the frequency for the time from zero to laptime is equal to the track length.

Errors can result from sound compression, the traction control can create wrong frequencies, spinning tyres give too high speed and in the braking zones a lockup can give you wrong frequencies.

Edit: Link corrected, and a big, big thanks to Reca for sharing this great post! =D>