How much faster are racing tires?

So a few weeks ago, I unknowingly got into a contretemps with Paul Hembery on Twitter.  Someone posted the question “how much slower would road tires be on an F1 car?”; Paul didn’t mention a time difference, so I commented on it and suggested what might happen to the tire from all the downforce and loading.

Sometimes I comment on people’s questions and answers to various people on Twitter.  After all, if they are replying in public, clearly they are leaving it open to comments from the public. I replied a different answer to what he replied, commenting on the propensity of road tires to chunk their shoulders when loaded excessively.  His reply?

@PaulHembery

“@malcolm33 it seems you know nothing about what you are talking about with the last comment. Intelligent opinions valid and welcome.

I was blocked after, so seeking clarification on his opinion was suddenly not possible.

After a minute, I started to think about why he might think have said that.  Perhaps it was because he viewed it as an attack on his beloved P-Zeroes, or maybe it was because we had a different set of assumptions governing our replies.  I believe he mentioned something about never getting heat into them and them lasting an incredibly long distance (he said “forever”, but clearly meant considerably longer than an F1 race distance).  I disagree with that.

I think I do know a bit of what I speak, since I also have a degree in mechanical engineering, have worked as a race engineer more than a few times and have 18 years racing experience as a driver (six years in karts, twelve years in cars – mostly GT, with a tiny bit of sports-racer and open wheel testing thrown in).  None of that has been in F1, or working directly with tire companies engineering their tires, but I definitely know how tires behave on the track, and how they can wear and degrade.

Firstly, how would a road tire react to an F1 car?  Assuming it was a Y-rated tire that could withstand the top-speeds, it shouldn’t explode at the end of a long straight.  I also doubt that tire-failure would occur in fast corners or heavy braking due to extremely high loads, but the tread pattern would definitely suffer from tread squirm.  Because of the tread squirm, the tire would heat up very quickly, likely overheating before the end of a single lap – it’s hard to compare to a road car that, while heavier, does not have near the downforce of an F1 car.  Another assumption of mine was that a brand new, unmodified tire was being compared; a trick often used in racing series that must use road tires is to shave off most of the tread, making the tread blocks considerably more rigid and less prone to squirm – this keeps heat down and improves performance.

If the road tire was shaved, I think it would still begin to overheat by the end of the lap, merely due to the intense energy being put through the tire from the downforce, the high g-forces and even the heat from the carbon brakes.  If it was not shaved, I would bet that by the end of the lap, most of the shoulders of the tires would have been chunked off, as the tire isn’t designed for 3+ lateral g-forces, considerable downforce and incredible braking from 300+ km/h.  The rubber used for a road tire is not designed for the tire temperatures than an F1 car would normally see; it would overheat at a lower temperature than a slick and drastically lose performance.

Secondly, how would the tire perform?  Clearly, a road tire has probably not been fitted to an F1 car for decades, and even then, it was likely only to move the car around.  One of very few possible comparisons would have to be the SCCA World Challenge; they have used street tires, semi-slicks (DOT-R, or “R-compound” tires), and currently full racing slicks.

2002 was the last year that they ran the standard “performance” road tires, Toyo’s T1-S model.  It was not a semi-slick or other racing-intended design; it was just a road tire.  After speaking to the drivers in the paddock, they mentioned that the tire would overheat in every session, causing the oils to come out of the tire, leaving a bluish colour on the surface of the tread and leaving the tire compound rock hard and virtually unusable.  Now I am not sure this would happen 10 years later with a P-Zero on an F1 car, but it might give a few clues.  David Farmer turned a fastest lap of 1:28.375 in the race at Mosport that year.

The next year, 2003, SCCA switched to the T1-R, which was a semi-slick, or “R-compound” tire.  This means it had a softer compound designed for higher operating temperatures, and it also had larger tread blocks for greater stability.  The tires worked much better, and allowed the cars to perform as they were engineered to, without frying the tires.  The fastest time at Mosport was by Bill Auberlen, at a 1:25.319.  That is a full three-second drop over an 88 second lap, which means if you were to fit a shaved street tire to Auberlen’s BMW, it would see a laptime increase of 3.6% (assuming the cars and driving are roughly equivalent from year to year).  David Farmer turned a 1:26.835, but finished much lower in the order, and therefore likely never got a clean lap during the race.

These tires were used from 2003 to 2010, after which Pirelli became the sponsor and wanted to use their slick tires.  From Racer.com, it was noted that there was a 3-4 second improvement from the semi-slick to the full slick by Pirelli: http://www.racer.com/world-challenge-teams-begin-pirelli-tire-tests/article/180898/ …  This test was done at High Plains Raceway, in Colorado, which would provide a roughly 2-minute laptime for GT cars.  At Mosport, this should equate to a 2-3 second laptime difference, or a 3.2% increase if you were to switch from slicks to semi-slicks.

Combining the two jumps yields a difference of roughly 6.8%.  This is comparing spec slicks from 2011 to road tires from 2002, and I assume that tire technology has improved since then.  For argument’s sake, lets assume that a road tire from today would be a little better, so the jump may only be 6%.  Shaving a street tire usually results in a 1-second advantage over a 90 second lap, which would be a 1.1% difference.  If the tire was as-delivered from the factory, that would bring our estimate up to 7.1%

Applying that to an F1 laptime, say around Silverstone (it was a British fan that asked, so I think it’s appropriate), Alonso’s pole time of 1:51.746 would have jumped up to a 1:59.780, a difference of 8.034 seconds.

My off-the-top-of-my-head assumption on Twitter was 10-12 seconds, and this calculation shows 8 seconds, assuming the tire does not degrade heavily or chunk during that single lap.  If the tire did degrade and chunk, the last sector of the lap could definitely see a decrease in performance, where a few tenths per corner could be lost.  A little over three tenths in six corners could definitely bump that up to a 10 second difference.

There are a lot of assumptions flying around here, relating different cars during different years, at a race that was moved from May to August, and then using those differences to apply to an F1 car with probably ten times the downforce, 20% more power, and easily half the weight.

So, what is the difference?  Probably 8-10 seconds at Silverstone, but until someone bolts a set of road tires on an F1 car, we won’t know for sure.

Hamilton’s Data…

Well, that was an interesting thing for Hamilton to tweet…

Things that I found interesting:

(I’ll refer to everything by the distance around the lap, which is noted at the bottom.  For example, Eau Rough is at 1200m or so)

1) Hamilton gains mostly under braking, rather than in the corners (300m, 2200m, 3000m, 6700m).  I would have expected the cornering speeds to be noticeably different, but they are actually quite similar.  Hamilton can just go that little bit deeper, and brake that much harder before the wheels lock, due to the extra downforce he was running.  With each steep drop in the speed trace (second trace from the top), you can see that Hamilton is just a little later on the brakes.

2) Hamilton messed up the third corner in the “Les Combes” section (corner 9, 2600m).  He has more downforce, so should be as fast or faster, but he must have had a moment there, as his speed drops mid-corner.  Unfortunately, the data is obscured by what seems to be the steering trace.  The slight correction of the steering seems to indicate that he understeered, as he only let up on the steering rather than going into opposite lock (either that, or he has superhuman reactions that corrected a slide so quickly that he didn’t need to get to opposite lock to save it… but I doubt it!).  You can see that in that short downhill run to Bruxelles (2500-2900m), the speed traces are parallel, so he isn’t losing time because of the wing – it was just his poor exit from the corner that lost him at least a tenth or so, where he should have gained at least one or two tenths.

3) Hamilton destroys Button under braking for the final chicane… only to lose most of that advantage by killing his corner exit (6900m).  While he was able to brake much harder (note the higher brake pressure he can apply without locking up, thanks to the added downforce – bottom trace, brake pressure overlaid with throttle position – 6600m), he probably ran wide mid-chicane, ruining his line on the exit.  Because of that, Button go the better exit and clawed back much of what he lost in the braking zone.

4) Through the easy-flat corners (Eau Rouge – 1200m; Blanchimont – 6200m), they both lose the same amount of speed.  Had this been a few years ago where Eau Rouge was almost flat, the data would have been much more interesting.  While Hamilton would have had more drag, he may have had as much as a 10-15 km/h advantage exiting Radillion or Blanchimont.  At some point, Button’s speed would eclipse Hamilton’s, but Hamilton could retain an advantage.  It’s counter-intuitive, but sometimes adding downforce increases your top speed down a straight, simply because you exited the previous corner that much faster – what you lose from drag is more than outweighed by what you gain from increased exit speed.  That’s why Le Mans cars are closer to medium downforce spec now, especially with the chicanes on the Mulsanne – the corner exits are very important.

5) Neither driver can trail-brake as hard into Bruxelles (2900m), due to the downhill nature of the corner shifting the balance forward, making the rear of the car “light” and twitchy.  The braking trace shows that as they turn in, they are braking with about half as much brake pressure as the entry to Pouhon (3800m); this could be partly due to the lower speeds and therefore lower downforce, but by watching the cars through that corner, some of it has to be because they are all quite twitchy on corner entry.

6) It is worth noting that at near top speeds, there is little-to-no brake modulation, as the car has so much downforce, giving the tires so much grip that arguably the best brakes in the world still can’t lock the wheels.  Note the braking into La Source (200m) – they are mashing the brakes, and then gradually easing off the brake all the way to the apex of the corner, mostly because they are losing downforce (and therefore grip) as they slow down. To avoid locking up, they must ease off the brakes as the limit of the tires gets lower and lower with the decreasing speed.

7) Both drivers seem to be quite smooth – a testament to the McLaren.  If you look at the whole lap, looking specifically at the steering trace (third trace from the top), there are very few corrections that were made.  Each steering input, Hamilton’s correction in Les Combes aside, it’s all very deliberate and consistent – no massive opposite lock moments chasing the car through the corner.  Then looking at the throttle trace, I can’t see anywhere where they had to lift to correct for any wheelspin – clearly the McLarens are putting the power down quite well.  It would be really interesting to compare to De La Rosa’s throttle trace, where I bet his steering and throttle inputs are far more erratic, for the simple reason that the HRT has less downforce, is probably twitchy in each corner, and is not able to put the power down nearly as well – therefore poor Pedro has to wrestle the car at the limit, rather than Jenson being able to finesse the car through each of Spa’s lovely sweeping corners.

Hamilton posted the photo because he was blown away by the differences between a high downforce wing and a skinny wing, likely magnified by his disappointment of being so far off the pace.  That was obvious to me (and probably anyone that understands the trade-off between downforce and drag), but it was the few other details that I found much more interesting.

F1 passing strategy with Pirelli tires

It was clearly evident in the Spanish GP this weekend that following another car would destroy your pace after only a few laps.  Why?  The lack of downforce caused the tires to overheat.  If there was a distinct pace advantage for the trailing driver, he could tail the leading driver for more laps, as the cornering speeds would be a little lower or the drop off in pace would be less than difference in potential lap time between the leading and trailing drivers.

Losing that aerodynamic grip just makes the car slide a little bit more, and the tires end up overheating and the driver losing major lap time.  Alonso was trying hard for a few laps to get close to Maldonado, but then he would lose grip and fall back. He would then take a few laps, cool his tires off, and try again really hard, finally killing his tires and almost falling back into Kimi’s grasp.  A similar thing happened with Hamilton, as he couldn’t get past Rosberg and ultimately could not defend against Kobayashi or Vettel.  On the contrary, Kobayashi divebombed whoever he came across in a matter of a few corners (or no more than a lap), thus never quite letting his tires degrade like Alonso or Hamilton did.

This new phenomenon will definitely make passing quickly a top priority, rather than taking the classic, measured approach of taking a few laps to assess the situation, find a weakness and exploit it in the most advantageous section of the circuit.  Now it’s a case of attacking while the tires still have grip to ensure that the wings get clean air, thus producing more downforce and maintaining tire performance.  If the trailing driver takes too long to pass, the tires will fall off massively due to overheating, and the leading driver will start pulling away.

This can also potentially bring about an interesting situation that could provide a stunning finish.  Theoretically, a driver could be held at bay just long enough that the tires are starting to overheat.  At just the moment when his tires fall off, he manages to pull off the pass; however, now he doesn’t have the advantage he did two laps ago, so the driver he just passed is equally poised to re-pass not just in the next corner, but perhaps even later in the lap, setting up a duel where the two cars could pass each other multiple times in a few laps.  I can only hope that we would see another Villeneuve vs. Arnoux type of battle, even if chances are slim.

Centre flap on Lotus’ beam wing

I noticed something about Lotus’ beam wing on Racecar Engineering’s excellent testing summary: http://www.racecar-engineering.com/articles/f1/f1-pre-season-testing-barcelona-1st-4th-march/

Note the flow under the beam wing, and how it converges to the middle:

Basically, it seems as if the centre flap, despite only being 20 cm wide, affects the entire span of the beam wing, even if only by a small amount. The low pressure region behind/below the flap pulls in, and therefore accelerates, the air under the entire beam wing, thus increasing downforce by itself, and also allowing the beam wing to have more camber while keeping the flow attached underneath.

Of course, like many details on an F1 car, it’s not *the* optimum approach, but the best approach they can take given the restrictive rules in place. A typical twin-element wing would be more efficient, but that’s not allowed; instead, they have to take advantage of the curious 20 cm free zone in the middle of the wing, since outside that zone there are no slots or flaps allowed. Just like the double-diffuser, high-noses, barge-boards and pre-2009 curvy wings, the little flap isn’t the best idea where there are no limits in that region, but it’s a brilliant work-around to a typically restrictive F1 rule.

Caterham comparisons… Why they’re doing well.

I think Caterham will do well this year – better than last year. A lot of people have compared them to new teams in the past, saying that other teams have achieved points-finishes and podiums much faster, but they seem to forget a lot of key differences.

1) Buying a team versus starting one from scratch – If you buy a team, you have all the key people in place, and you can evolve from a design that’s at least finished races. You don’t have to build everything from the ground up.  Red Bull got Jaguar as a starting point; Mercedes had Brawn, who had Honda, who had BAR, who had Tyrrell; even Ferrari had Alfa Romeo way, way back in the 1950′s (they even ran Lancia D50s when their own cars weren’t competitive against Mercedes!); the only “take over” that Caterham (nee Lotus Racing, then Team Lotus) did was to buy the Racing Technology Norfolk location and set up shop with a brand new cast of characters to run the team – hardly the same thing.

2) Starting a team with a limited budget – Toyota was the last team to start from scratch, but they rivalled Ferrari in terms of budget. It’s hard to compare Fernandes’ team to that of Toyota/TTE who were able to build up a facility that is now defunct for three years, yet is still the envy of most of the F1 field. Engine test benches that are second-to-none, and wind-tunnels that other teams are renting to verify their own results.

3) Starting a team in a no-testing era – Sure, you can replicate full races on the dyno, shake a car on a 7-post shaker-rig, play all you want in the wind-tunnel, but nothing beats pounding around a track for a few days.  Even Ferrari used to have their two wind-tunnels running 24 hours a day, 365 days a year, before the Resource Restriction Agreement. Caterham started their team and was thrust into their first race with only 12 days of testing. Contrast that to the Toyotas, Stewarts, etc, where they could test for a month straight if they wanted, and you can see the difference.  Want to test something minor today?  Slap on some hockey-puck demo tires, and go out for a “filming day” to make sure the car runs.  Want some real testing?  Need to wait until the Friday before a race – not the time when you want to try something revolutionary.

4) Starting a team in an era of unprecedented reliability – It’s crazy how many cars finish a race these days. If they don’t crash, they almost always finish. Engines are on a development freeze, so engine failures are few and far between. Minardi could pick up the odd points finish back when it was top-6 only; however, that was when half the field retiring was not overly uncommon. Finishing in the top-10 these days is probably more of an accomplishment than getting a top-6 back in 1990. I’d bet most drivers knew the access roads around each circuit back to the pits quite well – these days they’d get lost! Webber and Bourdais got their first points through high attrition. Those days are gone, even if only until 2014 when ERS becomes a major component of the drivetrain.

Given all of the differences these days that stack the odds against them, Caterham is doing well. They’re finishing lots of races, which is something that many other teams couldn’t achieve in their first few years. I think they’ll get their first points this year, and I hope they do. Despite the differences due to the modern high-reliability/no-testing era, you can only claim to be a “new” team for so long.

Front Wing F-Duct Idea

I recently saw this idea for a front wing F-Duct:

http://videos.caranddriverthef1.com/video/d0a182df002s.html

Not a bad idea, but it definitely has its flaws.

1. A major issue would be having the F-Duct in operation in all straight line situations, such as braking, when you clearly want lots of downforce.  This would limit braking ability.

2. You’d want downforce on both sides of the wing during cornering – it’s better to have more downforce on both sides, rather than less downforce overall albeit specifically focused on the inside.  Personally, I can’t see how trying to get more downforce solely on the inside would be much of a benefit; he could have had the idea that flattening the car to make the underbody work better, but I think that front wing is too flexible for it to work.  It’s mounted in such a way that any torque applied by having one side of the wing work harder than the other wouldn’t exactly transfer to the chassis, from what I can see, and would likely just twist the wing awkwardly.

3. How would you get it to transition between blowing the elements at a specific yaw to separating the airflow in a straight line?  What if the car were to slide just a little?  Suddenly the air goes from the outside passage (creating downforce) to the middle (reducing drag and downforce), thus causing the car to want to understeer… but once it understeers, the airflow goes back to the outside passage.  This could cause some instability…  Then, if the car were to twitch a little harder, it would go from the outside passage to the middle to the inside passage; in that time, the driver could be reacting to the aero balance when the car has less front downforce, and suddenly the front gets a lot more when it transitions to inside passage, creating a major shift in balance.  These rapid shifts in balance could make the car very difficult to drive, similar to the porpoising that Ayrton Senna experienced in early 1994 because his Williams’ front wing was too low.

4. The two functions (blowing for more downforce and separating for less drag) would likely require different amounts of air, and probably couldn’t use the same inlet.  For blowing, you need a lot of volumetric flow at a high velocity, which is something that that system would have a hard time with, considering all of the small inlet, small tubing, high number of bends and other restrictions the air would experience.  At best, you’d get a system that would barely offset the penalty induced by having a slot/disruption across the underside of the wing.  The drag reduction idea is plausible, but the blowing idea doesn’t seem to be.

5. He also has the slots in the wrong positions; you’d want to separate the flow on the main element of the wing, not the last flap… trip the flow earlier and you reduce that much more drag… in addition, you would want to blow the wing more toward the rear to add energy to the boundary layer – basically creating a jet that will help suck up the air toward the back of the wing to keep it attached.

I just can’t see how that would be a smooth transition, since by definition and design, it’s a logic system using fluid flow, so it’s either on or off.  With the old F-Ducts on the rear, it was driver-controlled (or speed-controlled in Mercedes’ case), so it would be up to the driver (or tuned to a specific speed) to engage, and not potentially engage mid-corner.

I think the best bet for this concept would be to use a Mercedes-style speed-controlled F-Duct for the straights only.  If I understand their system from 2010 correctly, it would engage only at speeds higher than the fastest corner that requires full downforce.  The same thing could be done to the front wing, ensuring that if the fastest corner is 215 km/h, then it would engage at 220, so every straight above that speed would get a drag reduction.  That way, in the corners, the yaw wouldn’t lead to rapid engagement and disengagement resulting in massive instability due to rapid shifts in aerodynamic balance.

Unfortunately, blowing the wing passively, using a ram-air intake, doesn’t seem feasible.  To get the airflow and the jet needed to measurably increase downforce, you would need a big scoop that would cause more drag than what it’s worth.  Personally, I’d like to see if he did more CFD than just the basic function of the fluidic switch, and see if he was able to come up with some downforce and drag figures.

Smart idea, but I don’t think his all-compassing solution would actually be feasible.

Edit: I might be wrong!  According to this site, Mercedes might be using this idea.

http://www.f1racingmotor.com/mercedes-wing-innovation-called-w-duct/

I still think they’d have to figure out how to have a smooth transition, and it might also have a speed-related engagement… and they’d have to figure out how to get a high-velocity flow from a small intake… and they’d have ensure that if the car began to slide, the aero-balance of the car wouldn’t have sudden changes.  Tricky!  …then again, if anyone can do it, it’d be an F1 team that’d have the ability.

Hello World!

This was a test post.

I like tests.

Especially when they have cool videos:

10 points if you can name someone that’s in the video.