Korean Grand Prix: Story From the Data

Posted on October 16, 2011


The Korean Grand Prix was a curious race from the point of view of the data. The pace of many drivers lacked consistency, especially in the first stint as the teams realised that the tyres were going to be fine after all and the track began to pick up grip. In fact, for intelligentF1 it was a very satisfying race, with the fears of multiple pitstops unfounded (as suggested here on Thursday) and a very competitive performance from the Toro Rossos (as predicted yesterday). But there are many questions – how quick were the Ferraris? How fast were the prime tyres? How much faster was Webber than Hamilton in the final stint?

The race opened in a flurry of activity, but quickly settled into a surprisingly slow pace. It took about eight laps for the drivers to gain confidence in the tyres, and then the lap times dropped by the best part of a second, for almost everyone. The first stint for the top six cars is shown in the chart below. The dashed lines are the intelligentF1 model fits of the underlying pace. The safety car laps are normalised to Vettel’s pace.

Vettel is about 0.4s faster than Hamilton, with Webber 0.6s down and the Ferraris 1.1s slower per lap. The change in pace at lap eight can be seen by the change in the slope of the traces – it is interesting to note that the increase in pace from the Ferraris is not as great, so it is likely that they were using their tyres harder in the early laps. This is supported by noting the undercut of the Ferraris by Jenson Button. The comfort with which he passed the Ferraris was made much bigger by the fact that the Ferraris were degrading their option tyres. The drift of their pace from the intelligentF1 curve fit is a sign that the tyres are entering the high degradation phase – both Ferraris lose a number of seconds between the stop of the chasing McLaren and their stops. That the Ferraris again (as in Singapore) remained on the prime tyres for the remainder of the race is likely to be related. This phase of the race was then reset by the safety car resulting from Vitaly Petrov’s assault of the Mercedes being driven by Michael Schumacher.

It was at this point that the effect of the prime tyres became evident at the front of the race. The first stint had demonstrated (Force India again) that the pace difference between the primes and the options was small, and at the front of the race the prime shod Webber chased down the option tyred Hamilton. The second stint is shown in the chart below. Vettel’s curve shows that he had a little in hand, but it is Webber who is 0.1s faster on the primes than he was on the options who brings the interest. The primes are new (worth about 0.3s per lap over used tyres) meaning that his pace is 0.4s up on his first stint. The step change in the pace of the Ferraris (lap 26) is seen on passing Rosberg, and the red cars are a stunning 0.5s quicker (+0.3s for new tryes) on the harder tyre than the tyre they struggled on in the first stint, and start closing in on the McLaren of Button.

The only change of position in the final stops was that Alonso jumped Massa. But there was something a little odd about this which can be seen in the data. The intelligentF1 fit shown in the chart below (red dashed line) is for Massa – not Alonso. From his pace in the second stint, this curve suggests that Massa should have had the pace to just about stay ahead of his team mate, but in reality he was a long way behind. This slow pace continued throughout the final stint. Fuel saving is a possibility, especially given that Rosberg was also curiously slow (by a similar amount) at the beginning of his final stint, and both Rosberg and Alonso ran dry before being able to return to the pits.

Either way, it got Alonso ahead, and he proceeded to chase down the McLarens, who were one of the few teams to be clearly slower on the prime tyre. Where Webber and Ferrari were faster on the prime, Hamilton was 0.4s slower and Button 0.3s. The pace advantage of Webber from the intelligentF1 model is shown by the blue dashed curve (with cross symbols) – it’s about 0.3s per lap. Hamilton did well to hold position in the final stint. Vettel? Well he was 0.8s slower on the prime tyre than on the option. Cruising? Certainly.

To the midfield, then and to the Toro Rossos. The first stint shows a number of key elements to the race, and is depicted in the chart below. Firstly, the two orange traces (Force India) are very close in pace (within 0.1s), but they are on different tyres. This showed that the prime tyre was competitive, and encouraged many to run it in the second stint. Secondly, as at the head of the race, there is little sign of significant degradation – both types of tyre were holding up well. The first four cars (two Mercedes, Petrov and Alguersuari) made an early claim to the points positions as their pace was clearly better than the remainder of the midfield, especially with Buemi stuck behind the Force Indias which Alguersuari had negotiated in the first half dozen laps.

Concentrating on the front half of this battle (once Petrov has accounted for himself and Schumacher), the second half of the race is represented by the chart below. Again the intelligentF1 model fits are the dashed lines. The pace of the Toro Rossos is clear, and Alguersuari struggled to get past Rosberg despite quite a clear pace advantage in the final stint. Rosberg’s stint is similar to some seen in other races with slow early laps paying dividends in higher pace later – this seems to be a characteristic of the Pirelli tyres, which will be investigated soon at intelligentF1. Di Resta was similarly caught by Buemi, but without the Merecedes straightline speed advantage, the Force India was unable to put up much resistance.

The back end of the midfield were joined by Lotus. They were within 0.5s of the Williams, the Saubers and Senna, and are easily plotted on the same chart. This is shown below with the intelligentF1 curve fits of Kovalainen and Perez.

The traces in the last stint have some strange features. Barrichello is totally consistent with the previous stints, and ended up ahead of Perez, who was clearly struggling with the tyres from surprisingly early in the stint, and the tyres gave up totally in the last two laps. Senna also had difficulties for the last half dozen laps, and was almost caught by the leading Lotus, who was already ahead of the three stopping Sauber of Kobayashi. Kovalainen was 43s from a point.

At the back of the race, it was a close run thing between the Virgins and Ricciardo, with Glock coming out on top, but the Australian overtaking D’Ambrosio in the second stint. The intelligentF1 model suggests that the HRT was about 0.5s per lap down on Glock, but it lost more time being lapped in the final stint, and was nearly caught by the slower D’Ambrosio. This is shown in the chart below.

In the end, the race was much more predictable than expected, and we did not have a day of chaos after all. Actually normal 2011 service was resumed – the tyres behaved well, DRS worked sensibly, there were great on track battles and different cars were fast at different stages of the race. And a young German guy called Vettel won.