Posted by: drracing | December 8, 2017

Suspension kinematics excel tool

Hi everybody,

this is just a short entry, to share a small project i am working on in the last few days. Since it is something i wanted to do since a long time, i really want to spend a few words about it here and, maybe, trace how it progresses.

As the title of this post reveals, i am working on a simple suspension kinematics tool in excel, which is able to calculate the position of each hardpoint, basing on the defined suspension movement and all the most important metrics.
For now i am only focusing on double wishbones, as this is the scheme which is mostly used on the race cars i am interested in, but it should be pretty easy to adapt it to be used for a generic five links geometry, in the future.

The method i used is based on the intersections of spheres in space. Each moving point of a suspension is derived as the intersection of three spheres, with the center of each seating on an hardpoint whose position is fixed or has been already identified in a previous step; the radius of the sphere is either a link length or the distance between the point we want to locate and another point of the same component (see for example a wishbone or the upright).
For the very first point (namely the outer point of the lower wishbone, in my case) we can use two close points (for example the two connection points of the same wishbone to the chassis) and a reference point, which can be placed everywhere and will be afterward used also as the input for each suspension movement. In my case, this point will only move in z (vertical direction), as this is all what we need to simulate how a suspension moves.

The tool can currently calculate the result of each heave and roll movement and of any combination of the two. The approach i used, mainly for convenience, is to assume that the chassis remains fix with respect to the global reference system, while the floor moves with respect to it, either translating, rotating or both.
Roll is assumed to be an equal and opposite movement of the two wheels of an axle, this avoiding any philosophical debate/headache about around which point/axis the chassis rotates.
Each metric is anyway always referred to the “new” ground position, so if for example i input a roll movement, all will be calculated with respect to the ground now seating with an angle with respect to the global reference system.

As the tool is still “work in progress”, it still doesn’t depict any of the components seating “above the control arms”, see for example the pushrod, rocker/bellcrank, spring/damper unit, antiroll bar and third elements. Also, for now i am not yet done with steering, but this is relatively easy to add as an horizontal movement of the steering rack and will most probably be the very next step.

This is how it looks like as of today:

Susp kin tool 1


Susp kin tool 2


Susp kin tool 3


As you may notice, the Roll Center is one of the metrics i identified, but there are also the instant centers (they are anyway necessary to determine Roll Center position) and, more importantly, the n-lines slope. The n-lines are the lines connecting the IC to the Contact patch and, effectively, the lines defining the only direction through which the suspension linkage can carry forces. The instant center can be really thought as the pivot about which the upright rotate around the chassis.
I thought and read a lot about anti roll effects recently and i am more and more convinced that the Roll Center is not really the best way to define how the suspension geometry contributes to the antiroll moment. Although the roll centers are pretty effective and relatively easy to use as a metric to just have a picture about the geometry’s antiroll contribution, if we really want to understand and describe what happens in a real cornering situation, they fall short, above all when they migrate significantly. That’s why n-lines look like a much more effective tool to understand what happens in certain situations. More on this will probably come in a future post.

For now, i simply enjoy my new tool!
I will publish an update to track the progress (hopefully) soon.

Posted by: drracing | November 28, 2017

WEC – Bahrain Race Analysis

Hi everybody!

Here we are again with the latest race analysis for this year, as after ELMS now also WEC 2017 season came to the end with a very interesting race in Bahrain.

Although the LMP1 title was already assigned in China, the race was anyway very interesting.
Toyota showed again a top form, above all with car n.8, who really dominated the event. Anyway, as we will see, the situation was a bit less biased on their side than in China in terms of performance, with Porsche who wanted to fare well the championship in the best way possible and who was not lucky in more than a situation.
Again, tyre degradation and strategy played a central role, both in LMP1 and LMP2.

Particularly exciting was anyway the battle in the LMP2 class, with the race victory deciding also the title winner. Both the crews fighting for the championship victory (Rebellion car n.31, who won in the end and DC Racing – Jota Sport car n.38) had some issues during the race, with the final stints of Bruno Senna slowed down by the power steering not working anymore and with car n.38 race in part defined by a refueling problem, not allowing them to complete the same number of laps per stint as their direct competitors and forcing them to stop once more than Rebellion.

The track itself is probably not as fascinating as other facilities around the world, but offers anyway some tricky points, like corner 9 and 10, where the cars have to brake while cornering and a couple of relatively fast sections (turns 6 and 7 and turn 12).



As we mentioned, Toyota n.8 dominated the event in Bahrain, leaving all the other contenders at least one lap down. This was Toyota’s third win in a row and fifth victory this year, meaning car n.8 won more races than anyone else in 2017.
Toyota’s car was extremely competitive at the beginning of the year and seemed to work extremely well again after the race in Texas, where Toyota was anyway not too far off from Porsche.
At the light of this strike of victories, it is really a pity that Le Mans proved to be so bitter for the Japanese Team, since that race not only is the most prestigious and well known of the WEC season, but also assigned double points.

In Bahrain, the crew of car n.8 did again everything right, with the car showing the best pace in the field and the team working the strategy perfectly.
Car n.8 was the only one who didn’t experience any issue or accident during the race and hence the only one doing only 6 pit stops, spending some 50 seconds less in the pit lane than car n.2, who finished second.


LMP1 - Pit Stops


In terms of performance, car n.8 was clearly the fastest car throughout the race. Although they didn’t sign the overall best lap (that went to car n.1), the following table (relative to the best laps and best 20, 50 and 100 lap times averages) clearly show their advantage on all the other competitors.


LMP1 - Best laps average table


Although car n.1 obtained the best lap in the race, the average of the best 20 lap times sees car n.8 already at the top, with the gap to all the other crews increasing as we look at more and more laps (average of the best 20, 50, 100 and “all clean laps” lap times).
Interestingly, car n.7 didn’t show the same pace and was faster than Porsche n.2, but slower than Porsche n.1 (who had an accident during the race, costing them time and a penalty). Although car n.8 could probably have been anyway faster than car n.7, it is worth to mention that Kobayashi had an accident about half way through the race; this produced a not well defined damage to the car that could have led a pace deterioration for car n.7, despite the reparation done in the garage.
The trends we just described are also easy to recognize if we take a look to the plots relative to the best 20, 50 and 100 lap times.


LMP1 - Best 20 laps


LMP1 - Best 50 laps


LMP1 - Best 100 laps


It is interesting to notice how, despite car n.1 being able to produce the best 10-11 lap times of the race, it completely falls behind car n.8 on the long run, with a pretty significant performance drop after the 12 mark.
Still, beside car n.8 being clearly the fastest one on the long distance, we cannot notice how car n.1 was still the one getting closer to car n.8 pace, with much better lap times than car n.7. Different story for car n.2, which was constantly out of pace.

If we look at how the lap times evolved during the race, we pretty much see the same trends in terms of relative performance of each car with respect to the others.
The plot below compares car n.8 to car n.1 and gives maybe also an indication about what could be a tendency in car n.8 to better preserve the tires than the Porsche.


LMp1 - All laps - 8 vs 1


Beside noticing that car n.8 line is nearly always slightly below car n.2 one, we can also identify a slightly bigger performance drop for the Porsche in certain stints, like for example the third one. It is anyway very hard to come to any clear conclusion, since there wasn’t apparently any dramatic performance degradation for any of the two cars.
The following two plots, referring to car n.2 and car n.7 performances compared to car n.8, also confirms how much quicker car n.8 was during the whole race and during some stints in particular.


LMp1 - All laps - 8 vs 2


LMp1 - All laps - 8 vs 7


As we mentioned already, particularly interesting was the pace difference between the two Toyota. Anyway, we have to always keep in mind that car n.7 had an accident with a GT Porsche at the beginning of lap 96, that costed them a puncture, a long time in the pit and, most probably, a damage to the car that was not completely repaired. It is not a case that, despite car n.7 being already slower than car n.8 during the third stint (which could be related to the typical Toyota split tyre change strategy: car n.7 changed the tyres during the first pit stop and was indeed pretty fast during the second stint, while car n.8 did it during the second pit stop and was quicker than car n.7 in the third stint), a bigger performance gap between the two cars seems to build up after the 100th lap, so immediately after car n.7 accident.

Let’s now break down our performance analysis considering each track sector and the relative performance of each car.
As usual, the track was divided in three sectors, as shown in the image below.


track map


The first sector is all about straight line speed and longitudinal acceleration, being composed by a big part of the main straight and three pretty slow corners, with two of them coming after intensive braking and followed by strong acceleration.
The second one contains a couple of relatively fast corners (turn 6 and 7) and the tricky corners 9 and 10, with the cars braking for turn 10 while negotiating turn 9. It also presents some intensive corner exit accelerations, above all the one after turn 10, which is followed by a pretty long straight. Here the hybrid deployment surely plays a role on the final performance.
The third and final sector has “only” three corners, but one of them is the fastest of the track (turn 12). Again, there are at least two corners exits where hybrid power deployment is very important, because leading to pretty long straight: turn n. 13 and turn n.14.

Sector one offers immediately the first interesting “talking point”, as car n.1 is clearly the fastest here, with car n.8 getting a bit closer only on the long distance.
This is indeed what the table below (relative to the average of sector 1 best 20, 50, 100 and “all clean laps” times) shows us, with car n.1 producing the best value for each one of the considered metrics.


LMP1 - Sec1 best times


As we mentioned, the gap between car n.1 and car n.8 gets a bit smaller as we analyze more and more laps. Car n.2 and car n.7 are relatively close to each other and both relatively far off the pace of their sister cars.
Looking to the plots relative to the best 20, 50 and 100 sector 1 times, we can get a better feeling about each car’s performance.


LMP1 - best 20 sec1


LMP1 - best 50 sec1


LMP1 - best 100 sec1


The plots give a better feeling not only about car n.1 performance advantage in sector 1, but also about car n.8 getting closer as we move toward the 100 mark. The gap between car n.8 and car n.1 reduces from about 0.2-0.3 seconds on the fastest lap of each car, to only about half of a tenth on the longer distance.
As we mentioned while describing the track, top speed and drag levels play surely a role in determining how fast a car can go through this first sector.
The trap speed data seems to confirm that having a better inline speed is important for sector 1 times, as car n.1 is constantly the one with the highest top speed.
Interestingly, car n.2 has very similar top speed to car n.8 but is significantly slower and this could indicate an advantage for Toyota during the acceleration phases (hybrid deployment).


LMP1 - Best 20 TS


LMP1 - Best 50 TS


LMP1 - Best 100 TS


The three top speed plots above (best 20, 50 and 100 top speeds achieved by each car), seems to confirm not only the advantage of Porsche n.1 on Toyota n.8 (the difference is about 4-5 km/h and we know this can also depends on how long each car coast at the end of the main straight), but also the difference between car n.1 and car n.2, which is pretty significant, while the two Toyota are closer together.
As we mentioned, interestingly car n.8 and car n.2 have pretty much the same top speeds.
As we saw already in other races, this seems to suggest, on one side, how the two cars of the same team may be running with different aerodynamic configurations and, on the other side, also how the Porsche seems to have less drag than the Toyota.

Sector 2 sees a completely different situation.
Toyota n.8 is definitely and constantly the fastest car, with a pretty significant margin on the closest competitor which is (no surprise) car n.7.
The table below, relative to the best laps and to the average of the best 20, best 50, best 100 and “all clean” sector 2 times gives already a pretty clear impression about the impressive pace of car n.8 compared to all the other crews.


LMP1 - Sec2 best times


Car n.8 is not only the fastest in sector 2, its advantage to the closest competitor (car n.7) also increases as we look at more and more laps, with more than 0.2 seconds gap on the “all clean” metrics.
It is also extremely interesting to see how far off Porsche’s pace was compared to Toyota’s one in this part of the circuit,  with a gap between car n.8 and car n.1 that moves between about 0.2 seconds for the average of the best 20 times, and more than 0.3 seconds for the “all clean” metrics. Again, it is important to notice how the gap between car n.8 and the other crews increases on the long distance, while the gap between car n.1 and car n.8 in the first sector showed exactly the opposite trend.
A picture tells more than thousand words; a plot is even better. So let’s take a look to the plots relative to the best 20, 50 and 100 sector 2 times of each car:


LMP1 - best 20 sec2


LMP1 - best 50 sec2


LMP1 - best 100 sec2


Above all by looking at the 100 best laps plot, one can really appreciate how much quicker than the other crews car n.8 was, in particular on the long run.
This plots helps also to see the performance drop that car n.7 had, maybe also because of the consequences of the accident and, in general, how much faster car n.8 was than the two Porsche.

This is also particularly important because, as we will see shortly, the performances of car n.8 and car n.1 in sector 3 were much closer to each other.
Already by looking at the table relative to the best average sector 3 times, we can see how, in more than a metrics, the difference in terms of pace between the two cars can be measures in terms of hundredths of a second, with car n.1 remaining always on top.


LMP1 - Sec3 best times


It is pretty impressive to see how close car n.1 and car n.8 were, above all when looking at the average of the best 50, 100 and “all clean” sector 3 times, with the latest showing a difference of only 12 thousandths of a second!
This well recognizable also by looking to the best 20, 50 and 100 sector times plots.


LMP1 - best 20 sec3


LMP1 - best 50 sec3


LMP1 - best 100 sec3


The first plot, relative to the best 20 sector 3 times, shows at best that car n.1 has actually the best sector 3 times overall. Anyway, if we look to the third plot, relative to the best 100 sector 3 times, we can really appreciate how the lines relative to Porsche n.1 and Toyota n.8 lie one on the other.
Car n.7 is constantly slower than all the other crews in this sector and this doesn’t seem to be linked to the accident, since we cannot really identify any point of the plot where their line lies below the ones of the other cars.
Sector 3 is indeed a mix of high speed straight and strong accelerations out of slow corners, which seems to require a combination of the strong points of each car that we already identified: Porsche has probably a top speed advantage on the penultimate straight, while Toyota could have a small edge on accelerations because of the hybrid deployment.

Combining the tendency of the three sector plots, we can come pretty easily to the trends shown by the best laps plots, with Porsche n.1 being faster up to the 11 mark or so and with Toyota n.8 becoming significantly faster afterward.

Closing with LMP1, this latest race was clearly dominated by car n.8, but it is always interesting to notice how each team interpret the rules and the car setup/configuration, producing better performances in certain tracks or section of a track and being a bit penalized in others. This was again the case in this race, with Toyota being so much stronger in sector 2 thus to allow them to compensate for a worse pace and sector 1 and a pretty much identical pace in sector 3, compared to the Porsche.
Hopefully, the new generation of LMP1 cars coming next year will offer the same technological interest and the same tension from a performance perspective.



The LMP2 class offered probably the most exciting show, with two of the cars contending the race victory also fighting for championship win. Rebellion’s car n.31 arrived in Bahrain with a tight advantage in the championship standings on DC Racing car n.38 and could, up to a point, managed the race. But the situation during the race changed many times, with both cars leading in different phases and with both the cars suffering some technical issues toward the end.
This had a direct impact on the pit stop strategy of car n.38, that was apparently not able to fill the tank completely in the last pit stops, this forcing them to progressively shorter stints and to a stop more than the direct competitors.
It must anyway be said that car n.38 also reacted very quickly to an announced FCY phase (because of a cat walking on the side of the track) that was then canceled and never started at all. The car anyway was already in the pit, with an early stop that could also have led to the one more pit stop at the end.


LMP2 - Pit Stops


The table above tells us that car n.38 spent about 35 seconds more in the pit lane than car n.31. If we consider that the gap between the two cars at the end of the race was about 10.7 seconds, we have something to think about!
Of course, car n.31 also had problems and could not push too hard at the end of the race because of a powering steering issue, with Bruno Senna offering probably one of the best performance of his life.
It is also worth to mention how car n.36, who finished fourth, used a different strategy, trying to triple stint the tyres on one side of the car, by changing only two of them during one of their early pit stops. This lead them to the shortest time in the pit among the cars we consider, but, as we will see, they also dropped down by several position during the stint immediately after this “ambitious” decision.
We cannot know exactly how much each crew pushed during the race, but the data we collected seemed to show that car n.38 had a small edge in terms of performance, although the difference between them and car n.31 was really small.
Our analysis will include more cars than usual, since I could identify more or less two “performance groups”, a faster one and a slightly slower one.
The following table, relative to the best and average of the best 20, 50, 100 lap times and “all clean” lap times, gives already an idea about each car performance during the race:


LMP2 - Best laps average table


Interestingly, the best lap overall has been obtained by car n.37, which was also very close to car n.31 up to the average of the best 20 laps metrics, but fell down on the long distance.
Beside the best lap overall, it is clear anyway that on the long run car n.38 had the best pace, with a gap on car n.31 (the closest contender) reducing a bit as we analyze more and more laps.
It is also worth to notice that car n.13 seemed closer to the sister Rebellion crew than in previous races, above all if we consider the average of the best 100 laps and the “all clean” laps.
Car n.26 was very competitive if we look at the best lap and to the average of the best 20 laps, but falls progressively down on the long distance.
Car n.36, despite its “alternative” strategy, had a good pace and was close to car n.31 in terms of performance up to the average of the best 50 laps metrics, but the gap beween the two cars performance increases a bit on the long distance.

The plots relative to the best 20, 50 and 100 lap times of each car help, as usual, to visualize better each crew’s pace relative to the others.


LMP2 - Best 20 laps


LMP2 - Best 50 laps


LMP2 - Best 100 laps


The first plot shows that car n.37 was indeed very fast if we only look at the best 3-4 laps and still pretty competitive if we consider the best 20 laps, but falls significantly down afterward.
Car n.38 is with no doubt the fastest on track, with car n.31 being close if we look at the best laps overall, falling a bit behind between the 5 and the 45-50 mark and then getting again very close on the long distance.
It is interesting to notice how, after the 50 mark, three cars (car n.31, car n.38 and car n.13) had pretty much the same pace, with only a small advantage for car n.38.
Car n.36 is a kind of special case in this race, partly because of the strategy the team adopted: they probably didn’t have the absolute pace of the best cars, as the first plot seems to suggest, but were anyway pretty close. Their line, in the second and third plots, lies pretty much on the car n.38 one between the 25 and 50 mark. After this, there is a significant performance drop and Signatech car finds itself between the bests in class and the “second group”, on the long run performance.

To better understand how the race evolved, let’s take a look at all the lap times of each car comparing them to car n.31 ones.
We start to the main contender, car n.38.


LMP2 - All laps - 31 vs 38


As we have seen, car n.38 was a bit faster than car n.31 and this is confirmed by the above plot. DC Racing’s crew is clearly faster than Rebellion’s one during the third, the fourth and the last two stints. We know anyway that Rebellion performance was compromised in the closing race phases by the power steering issue. We can also identify how the two cars stopped pretty much simultaneously the first three times, but this trend was broken starting with the fourth stop. From that stop on, car n.38 started coming into the pit lane always before and performed always shorter stints. The results was car 38 stopping once more than car n.31: this was probably one of the race’s deciding factors.

The comparison between car n.31 and car n.13 helps to confirm how close the two cars were in terms of pace, differently to what we have seen in other races.


LMP2 - All laps - 31 vs 13


More interesting is the comparison between car n.31 and car n.36, because it helps to highlight the impact of the strategy / tyres change decision that Signatech took, when they let Lapierre to triple stint the tires on one side of the car.


LMP2 - All laps - 31 vs 36


Although we all know the value of Lapierre as a driver, we cannot ignore the dramatic performance drop that the team experienced during the third stint, that also costed them many positions. This is particularly important considering that Signatech’s crew was in many phases as quick as Rebellion car n.31 if not quicker (like during the third, fifth and seventh stints). Of course, this strategy saved them some pit time, but the performance lost was too big to be compensated.

Let’s now break down also LMP2 teams performance, analyzing each track sector times.

Car n.38 was pretty much the fastest in sector 1, but the gap between DC Racing’s crew and Rebellion car n.31 is nearly nonexistent, above all if we consider more laps. The table relative to the best sector 1 time, the average of the best 20, 50 and 100 sector 1 times and of the “all clean” sector 1 times pretty much underlines how close to each others the performances of the two main contenders were.
Beside this, the most interesting note is how close car n.13 was to the the first two classified crews.


LMP2 - Best sec1 average table


Car n.31, car n.13 and car n.38 form the fastest group, while all the other cars are a bit slower and pretty much close to each other.
A look at the plots relative to the best 20, 50 and 100 laps helps to better understand the relative pace of each car compared to the others.


LMP2 - Best 20 sec1


LMP2 - Best 50 sec1


LMP2 - Best 100 sec1


As we may see, car n.38 line is constantly the lowest, if we only exclude the fastest sector 1 times of the race, up to about the 5 mark in the plots above.
Car n.13 and car n.31 are very close to each other, above all past the 19 mark.
Car n.36 is close, but still a small step behind the two Rebellion cars and DC Racing one, while the other three cars are pretty close together and a further step slower.
As we mentioned already during the LMP1 analysis, sector 1 includes the biggest part of the main straight, where top speed plays a very important role and two pretty hard brakings, leading to two slow corners that are followed by hard accelerations.
To have an indication about the reason why car n.38 is so fast in sector 1 and if there is any difference in the aerodynamic setup each team used, we can take a look to the top speeds data (the speed trap was placed at the end of the main straight).


LMP2 - Best 20 TS


LMP2 - Best 50 TS


LMP2 - Best 100 TS


As we would expect, car n.38 has the highest top speeds and this seems to (at least partially) explain why they were so quick in sector 1. They were probably running a setup with lower downforce/drag than other cars.
It is anyway interesting to notice that car n.31 has constantly lower top speeds than car n.38, still obtaining sector 1 times that are very close to car n.38 ones. It is also worth to notice that car n.13 and car n.31 seem to have similar performances in terms of top speed too.
This is not the case for car n.38 and car n.37, with the latter having significantly lower top speeds.
Car n.36, on the other hand, has slightly higher top speeds than car n.31, but is anyway a bit slower in the first sector, as we saw.

Sector 2 is more twisty and combines low and medium speed corners and a tricky braking point, where the cars brake while cornering.
By looking at the table relative to the best and average of the best times for this sector, we cannot really identify any dominant car, as nearly each metrics has a different crew signing the best result.


LMP2 - Best sec2 average table


It doesn’t comes as a surprise that some of the cars that had a lower top speed are pretty fast in sector 2, above all if we consider only up to the best 50 times: car n.37 signs the best sector 1 time overall and remains on top also if we consider the average of the best 20 times. On the average of the best 50 sector 2 times car n.37 remains on the same performance level of car n.38 and car n.31 but is beaten by car n.36, which was also one of the crews with lower top speeds and was close to car n.37 performance also in the average of the best 50 sector 1 times.
On the long distance, car n.31 and car n.38 return on top, followed closely by car n.13.

The following plots, relative to the best 20, 50 and 100 sector 1 times help us to visulazize the situation we have just described.


LMP2 - Best 20 sec2


LMP2 - Best 50 sec2


LMP2 - Best 100 sec2


Car n.37 is the fastest if we only look at the best 17 sector 2 times. Car n.36 took the lead afterward and remains the quickest up to the 44 mark. Past this point, car n.31, car n.13 and car n.38 come back on top and are clearly the fastest cars, all three with very similar sector times.

Sector 3 is again all about car n.38. Jota’s crew pretty much dominates in terms of performance in this track section, only missing the overall best sector time (anyway for less than 6 hundredths of a second) that was taken by the sister car n.37.


LMP2 - Best sec3 average table


The table above gives a first feeling also about how much faster car n.38 was compared to the direct championship competitor, car n.31. In each of the metrics we consider, the gap between the two crews is always slightly bigger than 0.1 seconds.
Car n.37 has the best overall sector 3 time, as we mentioned already, but falls behind on the long distance.
Car n.36 is relatively fast and pretty close to car n.31 performances, at least up to the average of the best 100 sector times.
Since each car sector 3 times are pretty much similar, the plots relative to the best 20, 50 and 100 sector 3 times are a better mean to get a feel of each team’s relative performance.


LMP2 - Best 20 sec3


LMP2 - Best 50 sec3


LMP2 - Best 100 sec3


These plots confirm that car n.38 is the only one really holding a gap on the rest of the crews, with its line lying constantly below all the others.
Car n.31 gets closer only on the very right part of the third plot, so namely if we consider the “slowest of the best 100 sector 3 times”.
The plot also helps, once again, to identify how extreme was car n.37 performance drop.
Interestingly, car n.26 is very close to car n.31 in this sector, if we consider the data relative to their best 45 sector 3 times.
Car n.36 seems to have potential but their performance also experience a significant deterioration past the 51 mark.

Closing, the analysis relative to the LMP2 class seems to show that, looking only at the performance side, race end result could have been more favorable to car n.38. Considering they stopped once more for refueling than the direct competitors, car n.31, we could ask ourselves how the race could have ended if they could run the same strategy.
On the other hand, we don’t have to forget that also car n.31 experienced some problems, above all during the closing phases of the race.
In any case it was an extremely tense battle, something really enjoyable for every fan. Rebellion succeeded in winning the first season of this new LMP2 era, which is even more remarkable if we think that they didn’t run in the (extremely competitive) LMP2 class last year and that they always showed extremely good performance, beside very very consistent race management.

Posted by: drracing | November 12, 2017

WEC – Shanghai Race Analysis

Hi everybody!

Here we are again with another race analysis. This time we will deal with the penultimate round of the World Endurance Championship, that took place in Shanghai last weekend.

Before we dig into the analysis, anyway, i want to share with you a video about one of the last simulation session we run, a few days before the Chinese race and on the Shanghai circuit.
The vehicle model in use is again the 2017 WEC Spec LMP2 and the simulations we did also helped to evaluate the effects of certain setup choices on final performance.
Here a link to the video.

I think it was all around a very exciting race, with a lot of interesting topics that can be analyzed and a lot of interesting battles/dramas happening on the track, above all in LMP2.

Tyres degradation has been, as always, an extremely important point, since the teams have to always double stint their tires to respect the given allocation. In Shanghai it seems it was even more important, because of the track being particularly demanding with this respect.
Also, the race was neutralized with a FCY only once, this giving even more room for proper wheel to wheel fights and with the teams having to take care even more of their hardware.

In the LMP1 class, the race was dominated by Toyota, who seems to have found much more pace than immediately after Le Mans and that has brought some very effective Aero updates in China. It could have been an 1-2, if car n.7 didn’t have an accident with a GT car nearly at the end of the race.

LMP2 was all about Rebellion’s car n.31, who took the pole in Qualifying and, excluding the pit stops phases, never lost the lead. Senna in particular seems to be living an amazing moment, with a form and performances that no other driver on the grid seems to be able to beat currently. Behind them, anyway, a lot of interesting battles took place, with car n.38 (DC Racing – Jota Sport) who had some tense moments and left China loosing the championship lead, now in the hands of Rebellion n.31 crew.

One very interesting point during the race was how several times LMP1 cars overtook LMP2 vehicles at the beginning of the long straight before the last two corners and then LMP2 cars overtook LMP1 cars again before the braking point, because of their significantly higher top speed (LMP1 cars were coasting at the end of the same straight).

Before we start to analyze the available data in details, i have to give a word of notice about this Shanghai race analysis. For a reason I ignore (I may suppose a problem with the time detecting system), a lot of the sector times (mainly relative to sector 2 and 3) were missing in the list I found in the official website. This means, although the trends we will identify should remain trust-able and realistic, the absolute values of the averages we will look at could well be wrong.
Unfortunately this is something that goes out of my control and that seemed to hit LMP1 data more than LMP2 ones.



As we mentioned, Toyota won with merit in China, their car working extremely well during the whole race both in terms of performance and tyre management. Also strategically, it looks like they did everything right.
If we take a look to the time each car spent in the pit, we see anyway that Porsche pit stops were actually slightly quicker, which give even more significance to the performance advantage shown by Toyota:


pit times


The table above actually indicates how all cars except car n.7 (who spent a long time in the garage for reparations after the accident with a Porsche GT car) did six pit stops; Porsche, anyway, spent less time in the pit, with the winning Toyota n.8 staying in the pit lane about two seconds longer than Porsche n.2 and nearly thirteen seconds more than Porsche n.1.
The advantage on pace was big enough to not only compensate these gaps but also to build a comfortable lead already short after the start.

It is no surprise to see this confirmed by the following table, listing for each car the best lap time achieved during the race, the average of the best 20 lap times, the average of the best 50 lap times, the average of the best 100 lap times and the “all clean laps average”:


LMP1 - Lap time average table


The best results for each metrics are written in red.
This table makes clear, on one side, that the two Toyota had indeed consistently more pace than the Porsche, with the gap between the two brands in each of the considered metrics staying constantly between 0.4 and 0.5 seconds, if we exclude the best laps and the “all clean laps” average.
Another interesting point is how Porsche n.2 was significantly slower than car n.1, while inside Toyota’s team, it is actually car n.7 that seemed to have the best pace, at least if we consider the best 20, 50 and 100 lap times. The gap between the two Japanese cars gets smaller and smaller as we move from the average of the best 20 laps to the “all clean” line and this is indeed what also the plots relative to the best 20, 50 and 100 lap times for each car tell us.


LMP1 - best 20


LMP1 - best 50



LMP1 - best 100


Some major takings, coming from these plots are:

  • car n.2 is by far the one with the worst pace throughout the race, but the question is if they really wanted to take any risk, knowing that they could bring home the championship win on race before the end of the championship
  • Porsche n.1 best lap times are indeed close to the ones of car n.8, but they get slower on the long distance
  • Car n.7 is clearly the fastest car, if we consider the best 20 and 50 laps plots
  • The two Toyota are very close to each other after the 45 mark, with car n.8 holding a small advantage
  • All cars seems to have a similar “gradient” (meaning with this how the lap times grow as we move from the best to the 100th best lap time) for their lap times curve with the exception of car n.8, with its curve staying pretty flat between the 15 and 60 mark, signalizing indeed how good Toyota’s performance was.


It is very interesting also to look immediately at the comparison between all cars’ lap times during the race, focusing in particular to car n.7 and car n.8:


LMP1 all laps - 8 vs 7

It is easy to see how actually during different phases of the races, one or the other car was quicker, with n.8 crew showing better pace during the first, third, fifth and seventh stint. Coincidentally, Toyota used a split strategy with their tyre changes, with one car changing tyres at a certain stop and the other at the following one. No surprise that each car has better performance than the other immediately after a tyre change, with car n.8 changing tyres at the second stop and then again at the forth and sixth (so before each stint where it had better pace than car n.7).

Porsche didn’t use the same approach and indeed the plot relative to the lap times throughout the race doesn’t show the “switching” feature we saw in Toyota’s one (simply car n.2 is most of the time slower than car n.1):


LMP1 all laps - 1 vs 2


Let’s now break down our analysis looking at each sector and at the speed trap data.
As often happens, the car was divided in three sectors, as shown in the following picture:


track map


The first sector includes a big part of the box straight and the first corner, which is a very tricky and long right turn where LMP cars brakes when already cornering, making it an extremely demanding and interesting point of the circuit. It is not uncommon to see different cars following completely different lines here. Beside the first corner (Turn 1 and 2 in the track map above), the first sector is composed by a second very slow left turn, followed immediately by a more open one that is seen by the drivers simply as a part of a single long bend. After a short full section section there is again an hairpin (turn 6) driven at very low speed. We can imagine, hybrid energy deployment plays here a very important role on performance.
Sector 2 is a very interesting mix of high and low speed corners: turn 7 is a very fast and long left one that lead to a slower right turn (turn 8), where the goal is to find the best line to prepare the braking for the following left one (turn 9). It leads into a relatively short straight through a second left bend (turn 10) which is run in full acceleration by LMP cars. Follows a pretty hard braking to turn 11, which is a pretty slow corner but leads into a very long right bend, where the car progressively increase their speed exiting in fourth gear and entering the longest straight of the track: this means, the exit speed of 13 is extremely important in defining final lap performance. Also in this sector, hybrid energy deployment is very important, but downforce also counts in defining how quick a car can go through turn 7 and 8.
The third and final sector is in big part composed by the longest straight of the track, where top speed and straight line acceleration play an extremely important role. The straight leads with a very hard braking into turn 14, again a very slow (right) hairpin. It is followed by a very short straight and the last corner (turn 16), a left bend which proved to be very tricky as many drivers were kept under investigation because of track limits infringements.

Let’s start with the first sector.
This is without questions Toyota’s hunting ground, with both car n.7 and car n.8 being clearly faster than both car n.1 and car n.2 and car n.7 showing a better pace than car n.8, above all if we consider the average of best 20 and 50 laps.
This is what the following table, relative to sector 1 best and average times, tells us, also underlining the small advantage of car n.7 on car n.8.


LMP1 - sec1 average table


As for the overall lap times, the gap between Toyota’s two cars reduce as we move from the best sector time to the average of the best 100 sector times and to the “all clean laps” average, with the two crews producing more or less the same results for these two metrics.
Porsche is again closer to Toyota if we consider only the best sector time, with car n.1, car n.2 and car n.8 obtaining more or less the same time, but looses ground as we move to the average of the best 20, 50, 100 and “all clean laps” sector times.
The situation appears even clearer if we look at the plots relative to the best 20, 50 and 100 sector 1 times:


LMP1 - sec1 - best 20


LMP1 - sec1 - best 50


LMP1 - sec1 - best 100

As we said, similarly to what we have seen analyzing the overall lap times, Toyota n.7 is faster than car n.8 if we look at the left portion of our plots, up to the 35 mark. On the right side of this point, car n.7 and car n.8 are indeed extremely close to each other.
It is also clear how both Porsche were 0.2 – 0.3 seconds slower than the two Toyota, with car n.1 performing a bit better than car n.2. Interestingly, anyway, the gap between the two Porsche remains constantly bigger than the one between the two Toyota.

As always, it is difficult to isolate the reason why Toyota was faster than Porsche in sector 1, as LMP1 cars are so technologically complex that this could also depend on more factors (aerodynamics, hybrid energy deployment, tyre management, etc). Looking to sector 2 numbers, it seems possible that downforce also played a role but this track is also particular demanding in terms of hybrid system recuperation and deployment and it would be no surprise if also on this side Toyota had an advantage

As we mentioned, sector 2 is (even more than sector 1) a sector where downforce can make a difference, since it includes two pretty fast corners (one at the beginning and one at the end of the sector, leading into the longest straight of the track) and one medium speed one. There are anyway also two very slow corner, with one in particular (turn 9) leading into an strong acceleration zone, where also the hybrid system can contribute significantly to the final performance.
And again, as for sector 1, in sector 2 the two Toyota were clearly faster than the two Porsche.


LMP1 - sec2 average table


The above table, relative to sector 2 best time and to sector 2 best 20, 50, 100 and “all clean laps” times average, see the two Toyota again on top, with car n.7 and car n.8 sharing equally the lead in each of the four metrics we consider. This time, anyway, the difference between the two crews is constantly smaller than in sector 1, with the gap between the best Toyota and best Porsche staying always between 0.2 and 0.3 seconds.
This time, the two Porsche were slower than both Toyota also if we look to the best sector 2 times overall.

The plots help again to visualize each car’s performance relative to the others.


LMP1 - sec2 - best 20


LMP1 - sec2 - best 50


LMP1 - sec2 - best 100


What these plot tell us is:

  • Both Toyota are faster than the Porsche.
  • Car n.7 show again a very high potential, above all if we consider the laps before the 35 mark.
  • After the 35 mark, car n.8 is a bit faster than car n.8, while in sector 1 they were more or less on the same performance level.
  • Porsche n.1 was sensibly faster than Porsche n.2, although car n.2 had actually a better sector 2 best time.


The summary of what we have seen in this first two sectors seems to suggest that Toyota could handle the more twisty / high longitudinal accelerations part of the track much better than Porsche.
This is particularly interesting thinking about the aerodynamic updates that Toyota used in China (and what could have been if they had them already before during the season), but it looks realistic to think that also on the hybrid side Toyota had some more flexibility.

Sector 3 anyway shows a pretty different situation. Sector 3 is in big part dominated by the long straight, followed by a very hard braking and a very slow hairpin. Top speed, hybrid performance and low drag play surely an important role in defining how fast a car is in this sector, but as the sector starts in a point where the cars are already traveling at high speed (exit of turn 13) and probably have already deployed a big part of their electric power, i suspect a lower drag can bring here a greater benefit.
If we look at the following table, showing the best sector 3 time and the averages of the best 20, 50 and 100 sector 3 times, we see for the first time a Porsche on top at least in some of the metrics:


LMP1 - sec3 average table


Porsche n.1 is indeed the fastest car if we consider the average of the best 50 and 100 sector 3 times, being anyway extremely close to Toyota n.7 also in the average of the best 20 sector 3 times.
It is also interesting to notice how also the performance of Porsche n.2 are pretty close to the fastest times in this sector, with n.2 crew taking the lead in the “all clean lap” average.
Porsche being here much more competitive than in the first two sectors is confirmed by the plots relative to the best 20, 50 and 100 sector 3 times:


LMP1 - sec3 - best 20


LMP1 - sec3 - best 50


LMP1 - sec3 - best 100


The first thing catching the eye here is that car n.8 is significantly slower not only than the sister car n.7, but also of both Porsche.
Toyota n.7 signs the best sector time and its performance is comparable to that of car n.1 up about to the 30 mark, but it drops a bit and also car n.2 becomes quicker on the long run.
Indeed Porsche were much more competitive in sector 3 than in the rest of the circuit.
Why this?
A small help can come from the Speed Trap data, as the TS was located at the end of this straight. As always, we have to be careful when we analyze this data for the LMP1 class, as the coasting strategies that each team uses can mix the cards a bit. Nonetheless, i think we can still get some important hints to understand each car performance:


LMP1 - TS - best 20


LMP1 - TS - best 50


LMP1 - TS - best 100


There are two things that we can immediately take, looking at these plots:

  • both Porsche were generally faster, in terms of top speed, than the two Toyotas
  • Toyota n.7 seems to be quicker than car n.8 up to the 40 mark and then drops down a bit. This seems to match also with the general performance trends of car n.7, who showed better pace than car n.8 above all to the 35-40 mark in each plot.

Top speeds analysis seems to suggest that the Porsche had either a lower drag setup than Toyota (which could also mean lower downforce, if the efficiency of Toyota really stepped up, because of the new updates; this seems to match with sector 1 and 2 performances), or a different hybrid deployment strategy, maybe prioritizing sector 3 and sacrificing a bit sector 1 and 2. Of course, also a mix of these two factors and, maybe, even others is possible. Maybe Toyota’s recovery strategy was also based on a longer recovery / coasting phase at the end of the straight. Really difficult to say.

Closing the analysis of the LMP1 class, China showed an extremely fit Toyota, most probably well helped by the small updates they installed on their cars for this race.
The Japanese cars were sensibly quicker than the Porsche and built their advantage mainly in sector 1 and 2, the most twisty ones and where downforce and hybrid energy deployment most probably play a bigger role.



The LMP2 class offered a very exciting race, with a lot of drama and wheel to wheel fights, involving also the main championship players, like DC Racing car n.38.
As all the teams use the same car (Oreca 07 – Gibson), there are less major performance differences compared to what we have seen for the LMP1 class (and in some previous LMP2 races), but we can still find some interesting points.

If we start considering the overall lap performance of the first four classified cars and of car n.28 (which came to sixth place, but was much faster than car n.25 who finished fifth), the following table immediately shows us that car n.31, who finished first, actually won on merit, not only because of a perfect strategy and because they avoided any mistake, but also because they had a very strong pace:


LMP2 - Lap time average table


Car n.31 was in each metrics in our table either the fastest car or very close to the fastest times. Beside signing the best overall lap time in the race, they are also on top if we consider the average of the best 20 laps and remain extremely close to the best performances also if we look at the average of the best 50 and 100 laps. They are again the quickest if we consider the “all clean laps” metrics.
Another interesting point, anyway, is car n.38 pace. Jota Sport’s crew was indeed extremely fast and, in terms of pure performance, could be beaten only by car n.31. Their race was anyway affected by a couple of mistakes/accidents that compromised completely their final result.
The plots relative to the best 20, 50 and 100 laps help us even more to understand how each team performed:


LMP2 - best 20


LMP2 - best 50


LMP2 - best 100


The first thing to notice, looking at the plots above, is how close all the cars we are analyzing were to each other, in terms of lap performance potential.
Car n.31 was indeed the fastest if we only look at the best 20 lap times plot. Above all their best 10 laps were sensibly faster than the ones of all the other cars.
Anyway, if we move our focus on the best 50 and 100 laps, we clearly see that the line seating below all the other was actually the one of car n.38, who clearly becomes the fastest car past the 10 mark.
It is also interesting to notice how close to each other were car n.31 and car n.36 (who finished second), in terms of performance, past the 25 mark.
Another interesting point is also that all the lines in our plot seem to show a very similar gradient from left to right (from fastest to slowest laps).

With this regard, it is very revealing to look also at how the performance of each car evolved during the race, also comparing each crew to the winning n.31.
If we first put car n.31 and car n.36 together, we see that the Rebellion crew was significantly faster than the Signatech one mainly during the second and the last stint, but also that the two cars were pretty close to each other (in terms of lap times) during the rest of the race. We could even go so far and say that car n.31 seemed to manage the tyres better, as they normally showed a slower increase of the average lap times during two stints (also in LMP2 each team had to double stint the tyres). But this could also be dependent on other factors, like traffic management.


LMP2 all laps - 31 vs 36


A comparison between car n.31 and car n.38 also offers some interesting discussion points.


LMP2 all laps - 31 vs 38

Car n.31 was again untouchable during the second stint, with Bruno Senna at the wheel, while Jota had his silver driver in the cockpit (Laurent), although Laurent is surely one of the biggest sensation of 2017 Season and is really fast.
Anyway, car n.38 produced in several stints a better performance than car n.31 and was sensibly slower only in the closing phases of the race.
This confirms their performance potential and that their race was actually ruined by unfortunate track action situations, but not because of bad pace.

Rebellion second car, n.13, was a step behind the sister crew in terms of performance and indeed one of the slower cars in our analysis (but we mentioned already how close these five cars were to each other in terms of performance), but still managed to close third, overtaking car n.38 in the final phases.
And indeed, if we compare the pace of car n.38 and car n.13 during the whole race, the first thing that catch our eyes is the dramatic performance drop that Jota’s crew had in the final two stints, while car n.13 was able to keep running with a much better performance.


LMP2 all laps - 13 vs 38


Let’s now try to break down each car performance by looking at the sector times.

In sector 1 car n.31 and car n.38 performance are very close, with all the other crews remaining anyway not too far, above all Signatech.


LMP2 - sec1 average table


Car n.31 and car n.38 metrics are always very similar, with the gap only slightly increasing as we consider more and more laps. Car n.36 follows very close behind.
A look at the best 20, 50 and 100 sector 1 times plots helps to get a better feeling about each team’s pace.


LMP2 - sec1 - best 20


LMP2 - sec1 - best 50


LMP2 - sec1 - best 100


Car n.31 seems indeed to have a small edge on the competition in this sector, with car n.38 being very close if we look at the 20 best sector 1 times plot, but loosing a bit of ground on the long run. Car n.13 is constantly slower than the other crews in this first part of the track.
Again, excluding maybe only car n.13, all the cars we consider are incredibly close to each other up to at least the 70 mark; this underlines not only the consequences of using the same chassis but also, most importantly, how good these teams work and how close they constantly are to the limit, after one season of development of their cars, setups and approaches.

Sector 2 offers a slightly different scenario, with car n.38 being constantly very fast and car n.31 not seating on top here, although remaining always among the fastest ones.
As we mentioned already, sector 2 is at least partially downforce dominated, but we better not jump to quick on conclusions and look first to the data of this sector and sector 3 afterward.


LMP2 - sec2 average table


The table above seems to confirm, without any doubt, that car n.38 was extremely competitive in this section of the track, although again the gaps between the cars in each metrics are extremely small.
This sector seems also to suite car n.13 pretty well, on the contrary of sector 1.
Is this also what the best 20, 50 and 100 laps plots tell us?


LMP2 - sec2 - best 20


LMP2 - sec2 - best 50


LMP2 - sec2 - best 100


Car n.38 and car n.13 look very strong, both if we look at the 20 laps plots or we concentrate on the long run.
Again, all the cars are pack extremely closed together and car n.31 is not far, although not shining as they did in sector 1 and in terms of overall performance.
Did they indeed run less downforce than car n.38?

This doesn’t seem what sector 3 data tell us.
In a sector where top speed (and hence low drag) play an important role, car n.38 and car n.31 are very close to each other in terms of performance and indeed have a slightly bigger gap to the other crews we consider, compared to what happened in sector 1 and 2.


LMP2 - sec3 average table


This is what the table relative to the best sector 3 times and to the average of the best 20, 50, 100 and “all clean laps” sector 3 times tell us.
Anyway, the plots relative to the best 20, 50 and 100 sector 3 times give in my opinion a better feeling about the performance situation in this section of the circuit.


LMP2 - sec3 - best 20


LMP2 - sec3 - best 50


LMP2 - sec3 - best 100


In all three plots it is crystal clear how car n.31 and car n.38 lines constantly lie below all the others, with car n.38 having maybe a small edge up to the 50 mark.
Car n.36, that had a very good overall pace during the whole lap, was indeed a step behind Jota’s and Rebellion’s crews in this sector, as was Rebellion second car n.13.

The top speeds data seemed to confirm that car n.31 and car n.38 were running similar drag levels:


LMP2 - TS - best 20


LMP2 - TS - best 50


LMP2 - TS - best 100


Car n.31 and car n.38 have pretty much always the best top speed and this is indeed no surprise, following the analysis of sector 3 times.
Car n.36 is, on the other hand, constantly the slowest one in terms of top speed, which seems to give a reason for their slightly slower times in sector 3 and their pretty good performances in sector 2.
All of this seems anyway to leave open the reasoning behind car n.38 being slightly faster than car n.31 in sector 2: this doesn’t seem to lie on different aerodynamic configurations.
Maybe the two teams were running different mechanical setups, which could also partially explain the differences in tyre management that we discussed before; but the reason of these differences could also simply “depends” on the driving styles of the drivers composing each crew or on the different race situations in which each team found itself during the race.
In general, car n.31 seemed to be a bit more consistent than the other crews (in particular car n.38, that was the closest in terms of performance), but they also had the (well deserved) advantage of being constantly in front, this giving a chance to manage the race instead of having to chase it, as it was the case for car n.38 during many phases; they most probably could not run in “save mode” and maybe had to stress its tyres much more (see for example the accident and relative spin that car n.38 had during the confrontation with G-Drive car, which surely didn’t do any good to the set of tyres they were using during that stint).

Posted by: drracing | November 10, 2017

2017 – 24 Hour Race Technology Article

Hi everybody!

This is just a very short post to share with you a couple of updates about my latest projects.

The most important news is that (not one but) two articles i wrote are featured in this year 24 Hour Race Technology issue.
The articles are about tyres and tyre modeling, with a specific focus on 2017 LMP2 tyres, their performance and the effects of specific parameters and strategy decisions on cars performance. There is a more theoretical part about tyre models and then a more “practical application” example, where my LMP2 vehicle model and my DIL simulator have been used to evaluate numerically the impact of said parameters on final vehicle performance and lap times. The results are then analyzed comparing logged data of different sessions.

Because of this article, i had a chance to come into contact with some very knowledgeable people, working respectively with Dunlop Motorsport, with bf1systems (a company producing very high level sensors, also for LMP1 and LMP2, including systems to monitor tyre pressure and temperature, including carcass temperature) and with a WEC LMP2 team. All of them provided extremely precious information and the WEC team also helped to validate the LMP2 vehicle model and its behavior (again, it was extremely interesting how good simulation can match reality).


mag 1


mag 2


mag 3


mag 4


You should definitely buy the magazine! Beside my pieces, there are also many other (more interesting) articles, including something written by Paolo Catone, father of the Peugeot 908 and, more recently, of the BR01 that raced last year in LMP2.

Beside this, i will start working with a new website, that will be online at the beginning of 2018 and will deal with news and topics related to sportscar, theRacingLine.
Hopefully the time at disposal will be enough to contribute substantially to this exciting project from my friend Dave Ellis.

Stay tuned as i will follow up soon with Shanghai WEC Race Analysis!

Posted by: drracing | October 27, 2017

ELMS – Portimao Race analysis

Hi everybody!

Last weekend offered another very interesting race, namely the ELMS finale in Portimao and we are here again with a race analysis to try to understand how each of the main characters in the show performed.

The race was pretty interesting, mainly because the title battle had still to be decided, although G-Drive had a pretty comfortable lead and only had to play safe to ensure the championship win. Anyway, we still had an exciting competition on track, with the car that won that started only in ninth position and had an issue with its fuel tank, which forced them to stop one time more than the other main competitors and with the car who lead most of the race (car n.32) and finished second in the championship standing having to serve a 55 seconds stop and go at the end of the race because of track limits infringements.

Beside this, the track itself offered a very nice scenario for LMP2 cars, with a lot of height changes, both relatively fast and slow corners and a very long straight, with pretty high top speeds. Beside this, the circuit is pretty bumpy and this poses a further challenge for the teams.

The race was won by Graff Racing, with their Oreca 07. Autosport Ligier JSP217 came to second while SMP Racing obtained another podium with their Dallara, closing in third position. For anybody interested in endurance racing and race engineering, it is always very nice to see such a variety, with three different chassis in the first three positions.
It was a pity that Dragon Speed, that started in second position, could not join the battle for the win because of an incident during the first few laps, that costed them a couple of laps.

Let’s start our analysis by looking at the time each team spent in the pit lane.


Portima PS


Car n.27 was the one spending less time in the pit, confirming the effectiveness that SMP pit crew showed already in the previous races. Anyway, the team opted not to stop during a FCY phase declared at lap 21 and that was used by nearly all the other teams to pit, saving some time at the end. On the other hand, it is interesting to see how SMP seems able to run longer on a full tank than other cars, most probably because they are able / brave enough to use their fuel to the very end (i doubt the car has a much lower fuel consumption than the competitors).
Car n.40 had to run one pit stop more because they apparently could not full the tank completely. This means, on the other end, they always run with a lighter car than the other crews at the start of each stint.
Car n.32 would have been also pretty good with their pit stop strategy, if they didn’t incur in the 55 seconds stop and go we already mentioned.
Car n.22 was not as efficient as SMP with their pit stops, but did only five stops, using the initial FCY at their advantage.
Car n.21 is included in this analysis as, even if closing the race only in ninth position, was indeed very quick, as we will see. The table above show pretty clearly how their race was completely decided by having to pit 7 times and by the issue they had at the beginning of the race, when Hedman went into contact with another car and lost two laps in the gravel.

Anyway, if we look at the pure performance of each car, Dragonspeed had indeed an amazing pace, at least with Lapierre and Hanley at the wheel.
The following table, showing the best lap, the average of the best 20 and 50 laps and the average of all clean laps (with the fastest time in each column shown in red), tells us that car n.21 was indeed the one with the best performance potential and actually the fastest as long as the two professional drivers were driving.


Best laps av


Dragonspeed Oreca is the fastest car if we consider the best lap and the average of the best 20 and 50 laps, with a gap of more than 0.5 seconds on the second fastest car on the single lap and about 0.3 seconds for the average of the best 20. The gap to car n.40 becomes anyway very small if we look at the average of the best 50 laps, with car n.21 falling inexorably down in the all clean laps average, where it occupies the last position among the cars we consider (again because of the heavy influence of their third driver).

Car n.40 is clearly the second fastest car, as we will also see shortly in our best 20 and 50 laps plots, but we have to take in mind also that car n.22 strategy was clearly to avoid any risk and surely they didn’t push to the limit during the race. During the closing phases, Hirakawa was actually coasting before each braking, maybe to save fuel and avoid the risk of having to stop and this confirms that the team was not really aiming at signing super quick laps but was simply controlling the situation.

The relative performance of each car compared to the others can be well visualized looking at the following plots, depicting the best 20 and 50 lap times of each crew.


Best 20 laps


Best 50 laps


Both plots suggest how car n.21 and car n.40 were clearly the fastest on track, with car n.21 edging the rivals although Graff’s car was often running with fuel less in its tank.
The only area of the above plots where car n.40 seems to be a bit quicker than car n.21 is on the very right side of the second graph.
Car n.32, which was contending G-Drive crew for the drivers title, was actually a bit slower than the Orecas of Graff and Dragonspeed, but seemed to be more consistent on the long run, with less performance degradation on the right side of the 50 best laps plot.
Car n.27 was more or less always off pace but still able to close third, helped by consistent drivers and, beside the strange decision of not pitting under FCY at the beginning of the race, by a very good strategy and quick pit stops.
Car n.22 was not really contending on the pace side, as the team didn’t really need it this time.

We can now look into more detail to each team lap performance, by analyzing the sector times.
The track was divided in three sectors, as usual, but on the contrary to many other races that we analyzed already, in Portimao there isn’t really a sector where a certain performance feature (like top speed or downforce or low speed handling) is dominating on the others, as each sector actually presents a good mix of low and medium-high speed corners, with sector one also including the main (and longest) straight.


track map


The times registered for sector 1 are inline with the whole lap performances, with car 21 being always the fastest as long as either Hanley or Lapierre were at the wheel. Dragonspeed’s crew advantage on Graff n.40 car is about 0.15 seconds on the best sector 1 time and stay at about 0.1 seconds if we consider the average of the best 20 and 50 sector 1 times. Car n.40 is on top if we consider the “all clean laps” average sector 1 time, with car n.21 following closely.
Car n.32 stays relatively closed to the best two cars, getting closer in the average of the best 50 sector 1 times or the “all clean laps” average.


Best sect 1 av


The plots relative to the best 20 and 50 sector 1 times gives even a clearer image of how car n.21 and car n.40 were in a class on their own in this track section, with the other three cars closer together but holding a sensible gap from the best two teams.
It is also interesting to notice how car n.21 line lies always below car n.40 one, if we exclude the laps following the 40 mark.
Among the “others”, car n.32 is the one staying closer to the two Orecas, while both car n.22 and car n.27 are sensibly slower.


Best 20 sec1


Best 50 sec1


Sector 1 analysis is particularly interesting also because the Speed Trap is located at the end of the main straight.
If we take a look to the top speeds, we see that, while car n.40 was, as in Spa, particularly fast, achieving the highest top speeds, car n.21 has actually pretty low top speeds, being actually also slightly slower of the sister car n.22.


Best 20 TS


Best 50 TS


This seems to suggest that, while part of the performance of car n.40 in sector 1 could depend on its top speed, car n.21 most probably was able to go quicker through turn 1 (a 4th gear corner), turn 2 (flat out), turn 3 (1st gear) and turn 4. This could mean that, once again, car n.40 went for a low drag setup, while car.21 worked more on the downforce and handling side. One thing to always keep in mind is also that car n.40 constantly had, at least at the beginning of each stint, less fuel on board than the other cars.
It is also interesting to notice how, while all the other teams had similar top speeds to each other, car n.40 seems to be on a different level with this regard.

Sector 2 showed a more balanced situation, with car n.21, car n.40 and car n.32 being closer to each other, in terms of performance, while car n.27 and car n.22 remain a bit behind.


Best sect 2 av


Autosport Ligier is always among the quickest, leading the group if we consider the best sector 2 time and the average of the best 20 sector 2 times. Anyway, it stays extremely close to the top also in the best 50 sector 2 times average and in the “all clean laps” average.
This seems to confirm the goodness of Ligier Chassis on the most twisty sections, with this sector being a mix of low speed corners (turn 6 and turn 9), full throttle bends and a pretty quick right-hander (turn 8), all combined with very suggestive height changes that make some spots blind.
The following plots, relative to the best 20 and 50 sector 2 times, give even a better feeling of how close to each other car n.21, car n.40 and car n.32 were.


Best 20 sec2


Best 50 sec2


Car n.27 is sensibly off pace in this sector, while car n.22 seems to have potential (if we look at the best 7-8 sector 2 times it had a pretty good pace) but is not particularly fast on the long run, the main reason for this being probably that they didn’t need to take any risk.

Sector 3 offers some interesting discussion points too.
As sector 2, also sector 3 is a good mix of slow and fast corners, with some very big height changes. It also include the very long and very quick last corner (turn 16), which puts a lot of load on the outside tires.
Car n.21 has again the best pace potential and signs the quickest time overall, remaining also on top if we consider the average of the best 20 and 50 best sector 3 times. As usual, its average pace drops dramatically if we consider the “all clean laps” average, because of the influence of their third driver’s performance.


Best sect 3 av


It is also interesting to notice how close to each other the two title contenders were in this sector, with pretty much the same result in the 20 and 50 best sector 3 times metrics.
Car n.27, on the other hand, gets very close to them on the long run, namely if we look at the “all clean laps” average.

Again, the plots give us a better feeling about the performance of each car relative to the others and they confirm car n.21 as the quickest, above all if we look at the best 20-30 sector 3 times, falling back a bit after the 30 mark.


Best 20 sec3


Best 50 sec3


Car n.40 is also very fast and follows car n.21 very closely, with its line actually lying below Dragonspeed one after the 30 mark.
It is interesting to notice how close to each other car n.32 and car n.22 are and how car n.27 gets on their performance levels after the 15th mark.
We could also speculate about the reason why car n.21 has a very strong pace if we consider the best 20-25 sector 3 times, but slows down a bit on the right side of the 25 mark. Since the car fell behind after the initial accident with Idec’s Ligier, a reason could be having to deal with traffic more than the other competitors, but this is really only a speculation. I am saying this, because they didn’t seem to have more tyre degradation than the other cars and were pretty much in line with car n.40 performance also at the end of the race.

We can get a feeling about this, by looking at the plot showing all race laps for each car. As usual, it is a very messy plot, but we can still extract some useful info out of it:


all laps all cars


  • car n.21 is extremely slow at the beginning of the race, with Hedman at the wheel
  • car n.21 is clearly the fastest car on track between lap 76 and 101, which are a big part of Ben Hanley’s stints and the beginning of Lapierre ones.
  • car n.32 was particularly competitive at the end of the race, compared to the other crews


If we analyze the plot comparing each car “race history” to the one of the winning team, we can extract even some more interesting information.
We start with car n.32.


all laps car 40-32

This plots shows again how car n.32 was slower than car n.40 pretty much for the whole race duration, with the exclusion of the last 15-20 laps.
This plot also shows how much closer to each other car n.40 pit stops had to be compared to the competition, because of the fuel tank issue and the additional pit stop of car n.32 because of the stop and go penalty.

Let’s now look at car n.27.


all laps car 40-27


As we saw already, car n.27 was in general slower and this is confirmed by its line lying the most of the time above car n.40 one.
We also identify SMP first pit stop happening out of the FCY window, around lap 29.
It is also interesting to notice how much car n.27 performance degrades between lap 40 and 55, before they stopped for their second pit stop: we could think they didn’t change the tires in their first stop and they suffered their degradation.
Another interesting point is that it looks like SMP crew was able to let their lap times to fluctuate less than the car n.40 ones, remaining pretty consistent on their pace, no matter how quick it was.

There is not too much to say about the comparison between car n.40 and car n.22, beside what we told already about the different pit stop strategy. Car n.22 was in general a bit slower, as most probably they crew never really “looked for trouble” during the race.


all laps car 40-22


Finally, a comparison between the two fastest cars, Dragonspeed and Graff.


all laps car 40-21


As we mentioned already when looking at all the cars together, the first very evident point is how much slower car n.21 was at the very beginning of the race, up to about lap 50. As soon as Ben Hanley jumped in the car, the pace increased drastically and get immediately close to car n.40 one. As we also mentioned, car n.21 was clearly the fastest machine on track between lap 76 and 101 and was very close to car n.40 pace in the last stint.
Indeed, car n.21 situation is always pretty interesting, as the potential it expresses is often extremely high, but is compromised by the non-professional driver who, most probably, is also the best business case for the team.
In general, among the three ELMS race we analyzed, car n.21 was the fastest car (or at least the car potentially faster than all the others) in two of them and was not on the best players level only in Spa.

Posted by: drracing | October 19, 2017

WEC – Fuji Race analysis

Hi everybody!

Here we are again after another WEC race, this time in Japan, at the beautiful circuit of Fuji.

Before starting the analysis itself, though, i would like to share again a video of a simulation session we did in preparation for the race, using the 2017 LMP2 vehicle model that was also employed for the ELMS – Spa race preparation video.
You find it here.
Unfortunately, we didn’t really have any chance of any comparison between simulation and reality, because the whole weekend has been affected by a very unpleasant weather, with rain and fog being always a constant theme in every session, in particular during the race.

The race itself has been wet conditions for its whole duration, with a lot of interruptions (in the form of red flags or safety cars) due mainly to scarce visibility. This has made the race more exciting, on one side, with strategy, vision and luck playing an important role on the final result and because it lead to many track fights, but also left (at least for me) some disappointment, since it was not possible to analyze how quick 2017 cars would really be on this very interesting track.
Also the distance covered was much shorter than it could be, with the final red flag signing the end of contentions after “only” 113 laps, with only 83 laps done under green flag.

These are the main reasons why also my analysis will be shorter than the previous ones. Nonetheless, there are a few interesting points that came out crunching the numbers publicly available.

The race was won by Toyota n.8 car, with the sister n.7 coming into second, in a very important 1-2 for the Japanese manufacturer, that has always obtained very good reults on its home track.
Car n.8 victory is even more important, considering that car n.2, which is leading the championship standing, finished in fourth place, plagued by a dramatic lack of pace.

Something similar happened also in LMP2, with Rebellion car n.31 taking the victory in front of an extremely quick Alpine n.36. DC Racing / Jota sport Oreca n.38 finished third and lost some of the advantage they had in the championship, with Rebellion car n.31 now closer.

Let’s take a look at how the performance of each of the main actors of this race was.


As we said, the very difficult conditions seemed to fit the Toyotas much better than the Porsches.
This is a very interesting point, above all if we consider that Porsche’s car n.2, the one who struggled the most in terms of performance during the race, actually signed the best lap time, with a gap of about 0.25 seconds on car n.8 best lap and was very fast at the beginning of the race.
If we also consider the average of the best 20 lap times anyway, Porsche n.2 seems still pretty competitive, being about three tenths quicker than car n.8 and with an even clearer gap to car n.1 and car n.7.
The situation changes completely, anyway, if we look at the best 50 lap times average: Toyota n.8 car is clearly the fastest, with about 0.3 seconds gap on car n.7 and car n.1, which are very close to each other. Porsche n.2 falls heavily behind and is nearly eight tenths slower than car n.8.


LMP1 average times table


The table above shows a summary where, for each average, the quickest time is shown in red.
We will not consider the “all clean laps” average this time, since the race was run for so long under safety car or yellow flag conditions, that it would not add much to our analysis.

The picture we got from the best and average best lap times is confirmed if we  look at the plots of the best 20 and 50 lap times of each car.


LMP1 Best 20 laps


LMP1 Best 50 laps


The first plot in particular (best 20 lap times of each car), shows how car n.2 is indeed the fastest if we only look at the best 13 lap times. Car n.8 follows closely, even in the very left part part of the plot.
The situation changes completely if we look at the second plot, anyway, where we can clearly see how dramatically the performance of car n.2 falls compared to all the other competitors (in particular compared to car n.8), above all if we focus on the right side of the 25 mark.
At the same time, this plot clearly shows how, for nearly the whole race length, car n.8 was consistently quicker the the other three crews, including the sister car n.7. It is also interesting to notice how close the performance of car n.7 and car n.1 was, with the line of each crew’s best 50 lap times intersecting the one of the other several times in this second plot. Car n.7, anyway, still has an edge on the very right side of the plot, confirming how also the second Porsche struggled in the last part of the race compared to the two Toyota.

According to the info i have and to what we learnt watching the race, Porsche (car n.2 in particular) had issues with the tyres, being affected by poor grip and by issues in bringing them up to temperature, above all in the second part of the race. 
The reason behind this seems to lie in the tyres specifications that each team planned/used for the race. As far as i know, there are two main types of rain tyres available for the teams, one fitting better heavier rain and lower temperature conditions (with Michelin being able to also modify them a bit on the track, if needed in the search for quicker heating up) and one adapting better for less extreme rain / less water / higher temperature conditions.
Apparently, Porsche could use the first tyre type at the beginning of the race but was unable to use them after the first pit stops; probably because of a different planning / pre race strategy, they didn’t have enough of these tyres for the race. As far as i know, these were also the tyres they used during free practice and qualyfing, where both Porsche were still extremely competitive.
If this is true, it means they found themselves running with tyres suiting the very special conditions that WEC encountered in Fuji much less than the ones that Toyota had at its disposal.
I think it is fair to say that, from a strategy perspective, Toyota did an excellent job in Fuji or, at least, any mistake.

According also to post-race interviews, Toyota n.8 was also helped a bit by a gamble with the fuel strategy, avoiding a third pit stop that actually both Toyota n.7 and Porsche n.1 did. As a consequence, car n.8 was also the one that spent less time in the pit.


LMP1 Pit times


It is now clear that car n.8 seemed to have a more consistent performance than the other cars. Let’s try to identify if this pace advantage was built on a specific part of the track more than in others.
As usual, the track was divided in three sectors, as shown in the image below:


Fuji track


The first sector is composed of about half of the main straight (which is very long) and the first corner, a very slow hairpin.
The second one includes the two fastest corners of the track, namely turn 3 and turn 4-5 and is (at least in dry condition) a very good section to evaluate of how much downforce a car has and, in general, how good the car’s balance and handling is in high speed corners.  There is then a mid-low speed corner (turn 6) that lead to a quick, full throttle section of the circuit (again, referring to dry conditions).
Finally, the third sector is composed mainly by slow corners, driven in first or second gear and, in some cases, with adverse camber producing very slippery car behavior.

Let’s take a look to Sector 1 times first.


LMP1 Sec1 average times table


The two Toyota are clearly in front here, no matter if we consider the best sector 1 time or the average of the best 20 and 50 sector 1 times.
This is particularly interesting, considering that, although the overall top speed was achieved by car n.7, the two Japanese cars seemed to have less straight line speed than the two Porsches, at least according to the trap speed measurements.
The best 20 and 50 sector 1 times plot confirms what shown by the table above:


LMP1 Best 20 sec1


LMP1 Best 50 sec1


Car n.7 and car n.8 are clearly quicker than the two Porsches, with car n.8 being faster than car n.7 up to the 46 mark and then dropping down a bit.
As we mentioned, this is particularly interesting because the two Porsche had often higher top speeds than the two Toyota, if we exclude for a moment the drop that car n.2 shows after the 25 mark and a big part of sector 1 is actually the final section of the main straight, where top speeds is surely important.


LMP1 Best 20 Top speed


LMP1 Best 50 Top speed

This drives some considerations.

  • we have to be careful, since i don’t know exactly where the TS was located: it was at the end of the main straight, but there is no indication about its exact distance from the start or the finish line. Maybe in certain phases of the race the Porsche n.2 had to brake so early (because of lack of grip) that the top speed measurement was affected (because the TS is in a zone where the car was already decelerating)?
  • For many laps there was a slow zone on the main straight, but i am not sure about how many laps exactly and if and how it affected cars speed
  • with LMP1 cars, there is always the unknown of how and where each car starts its coasting phase (energy recuperation strategy) before the driver hits the brakes
  • in general, it looks like Toyota simply had a better car than Porsche in this very special conditions and both crews were probably able to brake later and/or drive through turn 1 quicker and/or accelerate earlier at the exit of turn 1.
  • with such a bad visibility and the typical rain low grip conditions, i imagine that managing traffic situations was much more difficult than with a dry track, above all approaching such a hard braking zone as the one before turn 1. The car seating in front of the field (above all at the beginning of the race, when the lapping of slower classes has not begun yet), has sure an advantage.


As we are going to see analyzing sector 2, we could even think that the Toyotas had more downforce than the Porsche (and hence more drag too, which would explain the lower top speeds), but with a wet track one has always to be careful in coming to similar conclusions, as the pace of each car is dramatically affected by many factors, like the ability to have the tyres working in the right window (both in terms of temperature and pressure), mechanical settings (in order for the driver to have confidence but also to avoid typical rain issues like aquaplaning), the ability of each driver to deal with such a difficult scenario.

In any case, both Toyota cars seemed to have an edge on the two Porsche also in sector 2, where, in normal conditions, downforce would play a very important role.
Looking at the best sector times and average of the best 20 and 50 sector times, we see how car n.8 is always on top. Interestingly again, Porsche n.2 was extremely quick too if we look at the best sector time overall and also at the average of the best 20 sector 2 times (although in this later case the gap to car n.8 increases sensibly). On the other hand, we see again a dramatic drop when considering the average of the best 50 sector 2 times.
It is also interesting to notice how car n.7 and car n.1 have very similar performances no matter which value in the following table we look at:


LMP1 Sec2 average times table


Car n. 8 is pretty much in a league of its own, above all if we look at the best 50 sector 2 times average.
This is confirmed also by the plots relative to the best 20 and 50 sector 2 times.
Again, we see how car n.8 is constantly the fastest and how car n.2 had some good pace at the beginning of the race, producing very good times before the 16 mark, but falling down after.
It is also interesting to see how close the performance of car n.7 and car n.1 were, as the two cars indeed fight repeatedly with each other during the race.


LMP1 Best 20 sec2


LMP1 Best 50 sec2


On the contrary to what we have seen till now, sector 3 seems definitely Porsche’s territory, with both cars signing the best times here and with car n.2 being clearly the fastest if we look at the best 15-20 sector times.
The following table, relative to the best sector 3 times and the average of the best 20 and 50 sector 3 times underlines how Porsche had really an edge on the Toyotas in the final part of the track:


LMP1 Sec3 average times table


The picture appears even clearer if we look at the best 20 and 50 sector 3 times plot.
Interestingly, car n.8 was not particularly brilliant in this sector, being often also slower as car n.7.


LMP1 Best 20 sec3


LMP1 Best 50 sec3


Before closing with LMP1 and switching to LMP2, i would like to underline again how, beside a pure performance analysis, it is difficult to drive safe conclusions about each car performance in such difficult weather and track conditions, because a lot of different factors could play into the equation.
For sure, Toyota seemed to have an edge on Porsche and put in place a better strategy (with car n.8 also being helped a bit by the latest red flag) and, probably a better tyre choice/management.



Let’s now take a look at the LMP2 class. The race was won by Rebellion car n.31, followed by Signatech Alpine car n.36 and DC Racing / Jota sport car n.38.
Albeit all the teams using the same chassis, some differences could be identified in how each car performed and in how each team tackled both setup and strategy.

There has been some post race discussions because some teams didn’t have their Silver/Bronze driver in the car at all, because of the red flags and interruptions that reduced significantly the race time with respect to the planned six hours. But this goes beyond the scope of this article, that wants to focus more on cars/crews performance.

As we will see shortly, actually the fastest car on track didn’t win, but Rebellion managed the race very well and was also able to do one less stop than Signatech (who finished second) and than several other teams. The same is true also for Jota Sport car n.38. Interestingly, most of the teams stopped more or less at the same time for the first pit stop, but then differentiated their strategy and/or needed to stop because of driver changes (as happened to car n.36 just only five laps after its second pit stop).
Some drivers’ mistakes also affected how the race evolved: beside some more spectacular crashes, one effective game changer was probably the spin of Nicolas Lapierre at lap 51.

Because of all of this, it is no surprise to see that car n.31 and car n.38 were the ones spending less time in the pit lane, with car n.38 having the shortest pit time.


LMP2 Pit times


To start looking at cars performance, let’s consider the following table first, showing the best lap time and the average of the best 20 and 50 lap times of each of the first five classified cars.


LMP2 average times table


Already by looking to the above table, it is crystal clear that car n.36 was by far the fastest during the whole race. If the difference between car n.36 and car n.31 best lap is only slightly above 1 tenth, the gap increases progressively if we consider the average of the best 20 and 50 lap times, with a massive difference of nearly 0.6 seconds in the latter.

This is well confirmed by the plots relative to the best 20 and 50 lap times obtained during the race by each car.


LMP2 Best 20 laps


LMP2 Best 50 laps


Beside showing clearly how much quicker car n.36 was compared to the others, these two plots also confirms car n.31 being the second fastest car on track and car n.38 and car n.24 being very close to each other, as we also could see in the average times table.
Fourth classified car n.28 is constantly and sensibly off pace instead, but was one of the few crews that led his non-professional driver in the car for a pretty long time, during the first stints.

Let’s try now to break down each car performance analyzing each track sector times and the top speeds.

If we look at sector 1, we immediately notice how, actually, car n.31 is here quicker than car n.36, with all the first five classified crews being pretty much packed together with the exception again of car n.28, which seems to be a bit more off.
The table below, listing the best sector 1 time and the average of the best 20 and 50 sector 1 times, shows car n.36 being in front of car n.31 only if we consider the best time overall; car n.31 is anyway faster in both the best 20 and 50 sector times average:


LMP2 Sec1 average times table


This is pretty much confirmed by the plots relative to the best 20 and 50 sector 1 times of each car.


LMP2 Best 20 sec1


LMP2 Best 50 sec1


On the contrary to what we have seen in the LMP1 class, the information we gather here seems to match with what we see if we look at the top speeds of each car: as we said already, sector 1 is composed in a very big part by the main straight, therefore top speed plays surely an important role in defining how fast a car is in this track section.
Looking to the previous plots we would expect car n.31 achieving higher top speeds than car n.36, most probably because of a lower downforce/drag setup. And, indeed, this is exactly the case.


LMP2 Best 20 Top speed


LMP2 Best 50 Top speed


Car n.31 has the highest top speed in the bigger part of the best 50 top speeds each car achieved during the race. Car n.36 and car n.38 are both slower than car n.31 and that seems to match pretty well with what we saw looking at sector 1 times.
Interestingly, car n.24 is faster on the main straight than both car n.36 and car n.38, but its sector 1 times are generally slower than car n.36 and not dramatically quicker than car n.38.

Sector 2 seems to be the proof that our assumptions about each car aerodynamic setup could be correct, as both car n.36 and n.38 are faster than car n.31 here.
This is what we can conclude by looking at the table showing best sector 2 times and the averages of the best 20 and 50 sector 2 times and also what the graphs plotting the best 20 and 50 sector 2 times of each car underline.


LMP2 Sec2 average times table


LMP2 Best 20 sec2


LMP2 Best 50 sec2


Car n.36 is clearly the fastest in sector 2, with car n.38 following. Car n.31 and car n.24 are relatively close together.
It seems reasonable to conclude that car n.36 and car n.38 opted for an higher downforce/drag setup, while car car n.31 and car n.24 for a less draggy/higher top speed one. 

Sector 3 has a more balanced situation, in terms of performance, between the two contenders, with car n.31 and car n.36 being very close to each other and having a pretty big edge on all the other crews.
Car n.31 has the best sector 3 time overall, but car n.36 is slightly forward if we consider the average of the best 20 sector 3 times. Finally, car n.31 is again on top in the best 50 sector 3 times average metrics, but with the two cars practically achieving the same results (we have a difference of only 0,003 seconds!):


LMP2 Sec3 average times table


In a very twisty track section, where we mainly find slow corners and tricky road cambers, both the mechanical balance of the car and the ability of the drivers in the crew play a major role.
The equivalence of performance between the first two classified cars and their advantage on all the other competitors is reflected also in the best 20 and 50 sector 3 times plots:


LMP2 Best 20 sec3


LMP2 Best 50 sec3


Car n.36 has a small edge on car n.31 on the very left of both plots above, as confirmed also by the best 20 sector 3 times average. Indeed, the best 10-15 sector 3 times of the Signatech crew are slightly lower the ones of Rebellion’s car n.31.
But after the 15 mark the two cars seems to be pretty much a copy of each other.

Closing also this LMP2 analysis, it was interesting to find out how the winning car probably succeeded not because of a better pace, but because of a better strategy/management, while the fastest car was indeed the second classified one.
Also, it is always intriguing to notice how, although all teams use the same chassis, very often each one comes to a different setup choice and how this is reflected by each car’s performance in different track sections.

Posted by: drracing | September 29, 2017

ELMS – Spa Race analysis

Hi everybody!

I start to like this race analysis thing! That’s why i am here again with a new one, this one crunching the publicly available numbers relative to the ELMS race in Spa, held on the weekend of the 23rd and 24th of September.

Before to dive into the analysis though, i would like to share some data and a video about a simulation session done in preparation for the same race. The session took place on my simulator and the latest iteration of the LMP2 vehicle model has been used.
Here is a link to the video.
Incidentally, the best lap of the session was a 2’02.427, which compares pretty well to the pole position time of 2’02.457 done by Ben Hanley. In the video we managed to catch a lap slightly below 2’02.7.
This tells nothing in particular about the accuracy of the model, but as I had a chance to compare the simulations results with logged data, i can say i am pretty happy with how it matches to the real car!

And now, let’s look at Spa and the available data, to try and understand what happened exactly.

This beautiful track, which needs no introduction, offers always not only very exciting racing action but also a very interesting scenario from a technical perspective, mixing long straights and fast sections with quick and slow corners and leaving room for very different setup choices potentially producing a very similar lap performance.

The race was won with merit by Graff Racing car n.40, driven by James Allen, Richard Bradley and Gustavo Jacaman.
It was a complex race, with many neutralization phases either done with FCY or letting the safety car into the track. This was necessary, for example, when the Algarve car crashed at Pouhon and the car flipped over, after impacting with the barriers.
Because of the many neutralization periods, strategy played a very important role and teams went through the race using very different approaches, making it very difficult to read. For example, Dragonspeed cars (n.21 and n.22) pitted early at lap 8 when the first FCY was declared, while SMP, even if not using the FCY for an early pit stop, was able to run 19 laps with a full tank before stopping to refuel, while all the other main contenders stopped at lap 18.
To make all more spicy, the two leading cars both had to serve a drive through penalty at the end of the race and ended up with a gap to each other of less than 0.6 seconds.

Let’s look at the pit stops first and at the overall time spent in the pit lane.

Graff #40:                    5:58.598
G-Drive #22:               6:36.065
SMP #27:                     5:15.185
Autosport #32:           5:31.889
Dragonspeed #31:     6:51.258

The winning car n.40 spent 5:58.598 in the pit, including the late drive through, which took about 23.6 seconds and went into the pit lane six times.
Car n.22, who finished second, spent more time in the pit (6:36.065) and went in seven times, but, as we said, the first pit stop was done very early and under FCY, a condition where all the other cars are forced to drive slow anyway, while the latest was a drive through penalty that took 23.4 seconds. Without the drive through, the strategy could have proved very good indeed.
Car n.27 is the one who spent less time in the pit lane among the main contenders, with 5 pit stop and an overall 5:15.185.
Car n.32 also had a shorter pit time than the two leading teams, with five pit stops and an overall time of 5:31.889.
Finally, car n.21 also stopped seven times for a total of 6:51.258.
I am not totally sure about this, but my taking is that each team changed the tyres only twice during the race.
Again, it was a very complex race, from a strategy perspective. The five stop strategy that SMP and Autosport put in place was probably also in Graff plans, but the latest penalty forced them to stop once more. Interestingly, without that stop, they would have been in the pit for 5:34.987, not too much longer than what SMP did.

What about each car performance?

If we look at the best laps of each car, car n.22 is on top and the winning crew is actually “only” fourth:

Best lap times:
Graff #40:                    2:05.939
G-Drive #22:               2:05.236
SMP #27:                     2:06.25
Autosport #32:           2:05.719
Dragonspeed #31:     2:05.623

As we see, car n.22 was very clearly the quickest one on the single lap, with the good surprise of car n.32 being now much closer to the best contenders’ pace than in Paul Ricard.
Anyway, as we have seen already in the two previous race analysis, looking at the best lap time doesn’t tell too much in endurance racing.
If we take a look at the average of the best 20 laps, we start getting a better picture:

Best 20 lap times average:
Graff #40:                    2:07.158
G-Drive #22:               2:07.285
SMP #27:                     2:07.888
Autosport #32:           2:07.365
Dragonspeed #31:     2:07.912

The first thing we notice is car n.40 taking the lead, with car n.22 and n.32 being very close. It is not a case that also car n.32 was for a long time in contention for the lead and it is interesting to see how much closer to the Orecas they were here compared to Paul Ricard.
Another interesting thing here is that car n.21, which was clearly the fastest on track in Paul Ricard, was not as successful in Spa, at least in terms of pace. This is partially due also to how the race evolved, with FCY and SC periods probably affecting the stints of the two quick drivers in the crew, Lapierre and Hanley. One of the latest stints of Nico Lapierre was indeed very quick and car n.21 was probably the fastest car on track during that same stint. Still, the impression is that they struggled much more than in Paul Ricard, also in terms of pace.
All of these trends seems to be confirmed if we look at the best 50 lap times average, with car n.40, car n.22 and car n.32 ever closer to each other:

Best 50 lap times average:
Graff #40:                    2:08.32
G-Drive #22:               2:08.33
SMP #27:                     2:09.019
Autosport #32:           2:08.498
Dragonspeed #31:     2:09.327

It is impressive to see how car n.22 and car n.40 had nearly the same average time. As we will see later on, this is even more interesting considering that i suspect the two cars were running different setup philosophies.

Since the race was so much affected by FCY and SC situations, we will not have a chance to look at a best 100 lap times average (the leader only run 97 laps) and we will directly jump to the “all clean laps” times average (or something at least close to it, as several cars found themselves involved in a FCY or SC for a different time).

All clean laps lap time average:
Graff #40:                   2:09.059
G-Drive #22:               2:08.949
SMP #27:                     2:09.758
Autosport #32:           2:09.147
Dragonspeed #31:     2:09.94

Again, car n.40 and car n.22 are very close to each other, but with car n.22 now being slightly faster.
It is also very interesting to notice how car n.27 was pretty much always out of pace, compared to the other competitors, including car n.32, which was in every of the averages we looked at the third quickest.

If we look to the plots showing the best 20 and 50 lap times for each car, we pretty much identify the same trends. Car n.40 and car n.22 are pretty close together, with Graff’s car being clearly the fastest in the 20 best lap graph and the G-Drive one getting better if we look at the 50 best laps one.


Best 20 laps


Best 50 laps


It is also interesting to notice how car n.32 was pretty close to two Orecas pace for more or less the whole race and also was significantly quicker than the n.27 SMP Dallara. The latter, although being probably closer to the best cars compared to what achieved during qualifying, was still pretty much off in terms of lap times.
The plot also shows that car n.21 was not really as brilliant in Spa as it was in Paul Ricard.

Since the ELMS offers some variety in terms of Chassis used by each team, it is interesting to take a look at the sector times and analyse where each car was faster or slower, trying to find indications about each car strong or weak points and about how each team approached its setup.

The track was divided, as usual, in three sectors, as shown in the image below.


track map


The first sector is pretty much about top speed and having less drag. The second one is pretty long and represents a good test of both car’s low and high speed grip and handling. Downforce is very important here, together with a well balanced behavior in both low and high speed corners.
Last sector is again about speed and, partially, about braking stability (there is a very hard braking before the last chicane, where the cars downshift from 6th to 1st gear and from about 290 km/h to about 70).

If we look at the best sector 1 times, we can immediately recognize a very important trend:

Best sector 1 times average:
Graff #40:                   35.769
G-Drive #22:              35.981
SMP #27:                    36.038
Autosport #32:          36.086
Dragonspeed #31:    36.115

Just by looking at these numbers we immediately catch something extremely important: while all the other cars are very close to each other (with car n.27, car n.32 and car n.21 nearly all in less then a tenth of a second), car n.40 has a clear edge, being some two tenths faster than car n.22 and about three tenths faster than all the others.
This trend doesn’t change if we look at the average of the best 20 sector 1 times, but actually the advantage of car n.40 becomes even more evident.

Best 20 sector 1 times average:
Graff #40:                   35.984
G-Drive #22:              36.265
SMP #27:                    36.346
Autosport #32:          36.229
Dragonspeed #31:    36.335

Interestingly, the advantage of car n.40 on car n.22 increases, while the gap between car n. 40 and car 32 reduces slightly, with car n.32 being quicker than car n.22.
It is also interesting to notice how, in this first sector, the SMP Dallara doesn’t seem to be too far off from the competition, if we exclude car n.40.
If we look at the average of the best 50 sector 1 times, the trends remains very similar:

Best 50 sector 1 times average:
Graff #40:                   36.142
G-Drive #22:              36.430
SMP #27:                    36.554
Autosport #32:          36.374
Dragonspeed #31:    36.611

The gap between car n.32 and car n.40 reduces again a tiny bit but in general all the cars seem to increase their average times similarly compared to the 20 best sector 1 times average, with the exception of car n.27 and car n.31 that seem to have a slightly stronger performance deterioration.
This is particularly interesting for car n.27, as it is the only car in our analysis having only two silver drivers in the crew, both performing very similarly and, as far as we could see in Paul Ricard, pretty well too. We would expect a smaller effect due to the drivers when looking at performance deterioration in the 50 best sector 1 average than for other crews, but still the average time goes up slightly more than for other cars. This could signalize the car suffering more the effects of tyre deterioration in this sector, where, we have to remember, we also find the famous Eau Rouge corner, which could maybe become a bigger challenge if the tyres are not new anymore and if the car has less downforce than the competitors. These are of course mainly speculations, i am simply trying to interpret what we see, but of course i could be completely wrong.
In terms of performance deterioration, the worst case is still car n.21 (as in Paul Ricard), with the third (non-professional) driver probably still having a very important impact on the average performance of the crew.

If we finally look at the “all clean laps” sector 1 times average, we identify more or less the same trends:

All clean laps sector 1 average:
Graff #40:                   36.449
G-Drive #22:              36.761
SMP #27:                    36.940
Autosport #32:          36.739
Dragonspeed #31:    37.023

The main message, summarizing what we can deduce analyzing first sector performances, is that car n.40 seems to have a pretty clear edge on the competition. Graff’s car is consistently about three tenths quicker than the other cars, in a track section where performance are defined mainly by straight line speed.
This is well confirmed by the best 20 and 50 first sector times plots:


Best 20 sec 1


Best 50 sec 1

This seems to strongly suggest that car n.40 was running less downforce/drag than the other Orecas (and less drag than the other cars) and/or that they had more engine power. Since we deal with a spec engine class and the difference in time is pretty big, i think the reason should be mainly connected to a different aerodynamic setup, compared for example to car n.22 (although there is nothing like “all engines are the same” and i would not be surprise if, because of engine fatigue of other teams or simply because of the window that each manufacturer allows in terms of “power tolerances”, a part of a similar gap could rely on the engine itself).
Another interesting point that the plots confirm is also how competitive car n.32 was in this first sector, also compared to car n.22, which is consistently slightly slower.

Our conclusions seem to find further support if we look at the top speeds (the speed trap was located at the end of the Kemmel Straight, that means at the end of sector 1). Here, car n.40 has again a clear advantage on the competition, as we can see looking at the best speed overall and to the average of the best 20 and 50 top speeds each car achieved:

Best top speed:
Graff #40:                   301
G-Drive #22:              294.5
SMP #27:                    297.7
Autosport #32:          297.7
Dragonspeed #31:    292.9

Average best 20 top speeds:
Graff #40:                   297.95
G-Drive #22:              292.1
SMP #27:                    294.5
Autosport #32:          294.7
Dragonspeed #31:    290

Average best 50 top speeds:
Graff #40:                   296.5
G-Drive #22:              288.2
SMP #27:                    292.9
Autosport #32:          292.4
Dragonspeed #31:    287.7

As we could imagine, car n.40 is the fastest, with car n.27 and car n.32 following and being relatively close to each other and car n.22 being a bit slower.
Looking at the plots relative to the best 20 and 50 top speeds, we identify more or less the same situation but it looks like car n.27 had a slight advantage on car n.32 on the long distance:


Best 20 top speed


Best 50 top speed


Beside Graff having the best speed (and also the best sector 1 time, which looks very consistent) we see something interesting: although car n.27 seems to have a higher top speed than car n.32 on the long run, it constantly produces slower sector 1 times, which seems to suggest the SMP Dallara having slightly less drag than Autosport’s Ligier but also maybe some issues in driving fast through the Eau Rouge section, maybe because of lower downforce or different setup choices. The same conclusions seems realistic if we compare SMP Dallara to G-Drive Oreca, that has slower top speeds, but is nonetheless consistently faster in sector 1.

Contrary to sector 1, in sector 2 downforce and handling play a central role, while having low drag is surely not the top priority to obtain competitive times.
Following what we learnt analyzing sector 1, it is not a surprise to find out that the fastest car in sector 2 is actually G-Drive Oreca (n.22).
If we only look to the best sector two time of each car, the advantage of car n.22 on the competition is an astonishing half a second, with car n.32 and car n.21 following and car n.40 being the slowest.

Sector 2 best times:
Graff #40:                   57.588
G-Drive #22:              56.745
SMP #27:                    57.475
Autosport #32:          57.248
Dragonspeed #31:    57.249

The situation changes a bit if we look at the best 20 sector 2 times average, with car n.22 still being fastest than all the others, but with a much smaller gap, above all compared to car n.40 and car n.32, with the latter being very competitive in this second sector.

Sector 2 best 20 times average:
Graff #40:                   58.335
G-Drive #22:              58.157
SMP #27:                    58.566
Autosport #32:          58.241
Dragonspeed #31:    58.594

We see a similar picture if we also look at the average of the best 50 sector 2 times, only the gap between car n.22 and the closest among the others (car n.32 and car n.40) increasing a bit. SMP Dallara remains pretty much out of pace in this sector:

Sector 2 best 50 times average:
Graff #40:                   58.972
G-Drive #22:              58.717
SMP #27:                    59.197
Autosport #32:          58.966
Dragonspeed #31:    59.385

Nothing changes significantly in the “all clean laps” average, with car n.22 still on top of car n.32 and car n.40:

Sector 2 “all clean laps” times average:
Graff #40:                   59.651
G-Drive #22:              59.422
SMP #27:                    59.927
Autosport #32:          59.640
Dragonspeed #31:    60.267

The performance of each car compared to the others in sector 2 is somehow easier to perceive if we look at the best 20 and 50 sector 2 times plots:


Best 20 sec 2


Best 50 sec 2


Above all when looking at the best 50 sector 2 times plot, we clearly identify how big was the edge that G-Drive crew had on the competitors. We also clearly see that car n.32 and car n.40 performance were pretty close to each other. Again, if we assume that the two cars have the same engine power and we consider also what we saw in sector 1, this could indicate that the Oreca 07 has a higher aerodynamic efficiency than the Ligier Js P217 (in this case, in particular, meaning the Oreca had similar downforce but less drag), which seems to be reasonable considering that the Oreca 07 defined pretty much LMP2 performance target until today.

Sector 3 is, again, a quick section of the track where top speed plays a very important role. The only quick corner belonging to this sector is Blanchimont, which is an easy flat out with these cars. Beside this, as we said, there is a heavy braking into the bus stop chicane, which is composed of two very slow corners.
In this sector, car n.40 is again in front in terms of performance, above all compared to the direct competitor during the race, car n.22. The latter, gets closer to Graff’s pace only if we consider the last part of the best 50 sector times list: is it maybe connected to a bigger tyre wear affecting car n.40? Or maybe to a worse traffic management? Difficult to say.
Looking to the best sector 3 times, car n.40 is con top but interestingly the closest car following is car n.32:

Sector 3 best times:
Graff #40:                   32.175
G-Drive #22:              32.27
SMP #27:                    32.409
Autosport #32:          32.183
Dragonspeed #31:    32.259

Car n.40 gains a stronger gap on the competition in the average of the best 20 sector 3 times:

Sector 3 best 20 times average:
Graff #40:                   32.487
G-Drive #22:              32.608
SMP #27:                    32.618
Autosport #32:          32.615
Dragonspeed #31:    32.584

Car n.22, car n.27 and car n.32 are extremely close to each other, with car n.21 being slightly quicker than the other two.
The situation changes a bit if we look at the average of the best 50 sector 3 times, with car n.40 being pretty much on the same level as car n.22 and car n.27:

Sector 3 best 50 times average:
Graff #40:                   32.747
G-Drive #22:              32.795
SMP #27:                    32.797
Autosport #32:          32.894
Dragonspeed #31:    32.896

It is interesting to notice how close to each other are car n.22 and car n.27, confirming in this sector the Dallara is pretty competitive.
All cars are also very much packed together if we look at the “all clean laps” average, with car n.40 still holding a small advantage and car n.22, car n.27 and car n.32 being very close to each other.

Sector 3 “all clean laps” times average:
Graff #40:                   33.06
G-Drive #22:              33.12
SMP #27:                    33.178
Autosport #32:          33.153
Dragonspeed #31:    33.261

As always, the best 20 and 50 laps plots give a more “visual” feeling about the relative performance of each car.


Best 20 sec 3


Best 50 sec 3


Up to the 30 mark, Graff’s Oreca is clearly the fastest car, while between the 30 and 50 mark the performances of car n.40, car n.22 and car are very close to each other.
Car n.32 is very quick on the very left side of the plot, but tends to loose performance and, in general, seems a bit slower than the competition, despite having average times that are not too far from the ones of the other cars.
Again, it is also interesting to notice how in this sector the n.27 SMP Dallara pretty much matches car n.22 pace, in a track section which is probably the one resembling more closely the second sector of Paul Ricard track, where car n.27 was very competitive (mainly inline speed and a hard braking).
Interestingly, it looks like Autosport’s Ligier performs better in the first sector than in the third one, although in both of them top speed plays a central role. We cannot forget, anyway, how an important part of first sector performance depends on how good the car flies through the Eau Rouge complex.

To close we can take a quick look at the lap time each car produced during the race at each lap. Again, the plot comparing all cars is pretty messy (even more in this race, because of the many SC and FCY that came in), but still tells interesting things:


all laps all cars


Some points we can extract just by looking at all the lines are:

  • Car n.21 lost a lot of time in the first two stints, where the non-professional driver was at the wheel
  • Car n.21 was extremely quick between laps 60 and 70, but was in general not as quick as it was in Paul Ricard compared to the competition
  • Car n.40 had a pretty bad performance between laps 60 and 70, being in this interval of time slower than car n.22 and car n.32
  • maybe also because of the many safety cars, we don’t seem to identify a dramatic performance deterioration as, for example, in COTA WEC race


More information can be deduced by looking at the comparison between car n.40 and each other car.
The following plot compares car n.40 with its main contender for the final victory, car n.22.


all laps 40 vs 22


Car n.22 was apparently slightly faster during the first stints and again between laps 63 and 76, but slower in the stint immediately after and in the final stint, when it probably got caught in traffic. Car n.40 was faster also between lap 45 and 52.

What about SMP Dallara?


all laps 40 vs 27


As we have already seen, car n.27 was constantly a bit slower than car n.40 and never really on the same pace as the Orecas. The green line in the plot above lies nearly always above the blue one, in particular during the last four stints.

A comparison between Graff’s car n.40 and Autosport’s car n.32 is, on the other hand, a bit harder to read:


all laps 40 vs 32


Car n.32 was indeed very fast in the first part of the race and again faster than car n.40 between laps 65 and 75, as we saw. Anyway, it didn’t match car n.40 pace in the last stints nor it did between lap 45 and 52.

Finally, as we already had a chance to say, car n.21 was not so fast in Spa and this is again confirmed if we look at the following plot:


all laps 40 vs 21


Until lap 60, car n.40 seems constantly faster. Car n.21 runs some very competitive laps 61 and 74, but also afterward it never really match car n.40 pace anymore, also showing maybe the only case of visible performance degradation, with its lap times constantly increasing in the last two stints.

Closing, this analysis confirms that car n.40 not only worked out its strategy very well, but also obtained its victory relying on its pace, which proved to be extremely good, also thanks to some setup choices that were apparently different than the other Orecas (see for example Dragonspeed – G-Drive).
Spa is surely one of the best track for the Oreca to express its potential, but it was also interesting to see how the Autosport’s Ligier was not too far off from Orecas pace (above all car n.22), while the Dallara seemed to struggle more.

Posted by: drracing | September 24, 2017

WEC – COTA Race analysis

Hi everyone!
This post will be again an attempt at an Race analysis, as i did in my previous one. This time we will look at WEC COTA race data.
Once again, i hope i will not come to completely wrong conclusions, trying to understand the how and why of this race (in terms of performance) by just looking at the numbers openly available online

Let’s start by saying, it was a very interesting race, with a pretty close battle (at least up to a point) in LMP1 and with very nice discussion points in LMP2 too (although having in the WEC only one chassis maker, there is some less variety, also in terms of performance, compared to the ELMS).
The race had only one safety car neutralization, with this being made not with a full course yellow, but letting the safety car into the track, practically deleting any existing gap for many of the track battles that were taking place.

Temperatures were very high and this has surely been an important element, also in terms of tyre degradation.
LMP1 teams also always had to double stint their tyres (i am not sure if LMP2 teams also did it, unfortunately i could not take note of every pit stop) and Toyota and Porsche did that in different moments during the race (Toyota double stinted its first set of tires, while Porsche did change the tyres at the second stop, following always alternating one stop with full service and one only with fuel and driver change only). That lead to interesting situations on track too, since the two marques often had a different tyre status in a certain phase of the race.

We will look at LMP1 first and then move to LMP2.


The race has been won by Porsche, who also put in place team orders to favor their championship leading car (Car n.2), despite being car n.1 most of the race slightly quicker.
The good surprise was that Toyota, after having shown no promising pace in qualifying, was pretty much close to Porsche during the race, with very exciting attrition taking place on track between car n.1 and car n.7 in a fight for second place.
During the race, car n.8 also suffered a small issue with the hybrid system, apparently becoming unable to play with the boost “freely” when needed (for example for an overtake maneuver) but still managed to end up in front of the sister car.

Looking at the fastest laps first, Porsche was slightly ahead, but with a much smaller gap than what we saw in qualyfing:

Porsche #1:      1’47.149
Porsche #2:      1’47.302
Toyota #7:        1’47.391
Toyota #8:        1’47.556

The gap between the two big teams stays more or less the same also if we look at the average of the best 20 laps:

Porsche #1:      1’47.768
Porsche #2:      1’47.908
Toyota #7:        1’48.022
Toyota #8:        1’48.011

Car n.1 is still the quickest and this was a topic of discussion during the race, since its performance was clearly sacrificed to favor car n.2, currently leading the world championship. The two Toyotas are extremely close together, but the situation changes already a bit if we look at the average of the best 50 laps:

Porsche #1:      1’48.135
Porsche #2:      1’48.279
Toyota #7:        1’48.353
Toyota #8:        1’48.470

As we can see, the gap between Toyota and Porsche becomes slightly smaller and the Toyota n.7 seems to now show more pace than the sister car. It is still amazing to see how close each car is to the others, to underline (if necessary) how good the drivers are and how intense is the competition.
If we consider now the average of the best 100 laps, we start to see the effects of team orders inside Porsche, with car n.1 and car n.2 being extremely close.
As a consequence, the gap between Porsche and Toyota becomes even smaller than before:

Porsche #1:      1’48.630
Porsche #2:      1’48.681
Toyota #7:        1’48.763
Toyota #8:        1’48.875

Interestingly, Toyota n.7 car is still faster than car n.8.
If we finally look at the “all clean laps” average, we still see a similar situation but with the gap between Toyota and Porsche now stretching again a tiny bit:

Porsche #1:      1’49.416
Porsche #2:      1’48.485
Toyota #7:        1’49.568
Toyota #8:        1’49.514

Car n.1 is still the fastest, while car n.8 becomes faster than car n.7.

Just looking at the numbers, we can identify how, on one side, Porsche had a small advantage on Toyota (above all considering car n.1) in terms of pace and how car n.1 was clearly the quickest.
This seems to be confirmed also by the plots looking at the best 20, 50 and 100 laps.


LMP1 best 20 laps


LMP1 best 50 laps


LMP1 best 100 laps


We immediately see, once again, how car n.1 has the best pace, with the only exception being in the window after the 65th best lap, where it becomes slightly slower than car n.2.
if we look at the best 20 laps plot only, the two Toyota seem to be very close to each other; but pretty big gap between the two emerges after the 22nd best lap mark, with car n.8 becoming sensibly slower than car n.7, at least up to the 90th best lap mark.

Clearly, in this second phase of the season Porsche has an advantage on Toyota. It can be anyway interesting to try to understand if there are some track sections where the gap is bigger and we will try to do this by looking to sector times.
The track was divided in three sectors, as shown in the following image:


track map


A first relatively short first sector is followed by two longer ones. Last sector is a pretty twisty one, with several slow corners and a very long right double bend (turn 16, 17 and 18) where both tyres and aerodynamics are very important, followed by a medium speed left corner (turn 19).

Analyzing the first sector, if we would just look at the best times of each car, we could come to partially wrong conclusions. The two Porsche are clearly in front, but the situation changes slightly when looking at the average of the best 20, 50 and 100 laps.

First Sector best times:

Porsche #1:      22.567
Porsche #2:      22.737
Toyota #7:        22.891
Toyota #8:        22.936

First sector best 20 times average:

Porsche #1:      22.962
Porsche #2:      22.968
Toyota #7:        22.960
Toyota #8:        23.007

First sector best 50 times average:

Porsche #1:      23.046
Porsche #2:      23.05
Toyota #7:        23.004
Toyota #8:        23.047

First sector best 100 times average:

Porsche #1:      23.121
Porsche #2:      23.133
Toyota #7:        23.061
Toyota #8:        23.099


We immediately see how, actually, all the car have pretty similar performances, also because we are dealing with a very short sector. We also notice how the Toyotas stay more constant, compared to the Porsches and become slightly quicker than the competitors already in the best 20 laps average (car n.7).
Since the first sector is mainly composed by the box straight and the first corner, could it maybe lie in the hybrid boost strategy/flexibility? We are of course just speculating, the cars could also have different aerodynamic properties, or different traction in corner exit, just to name a few possible reasons.

Interestingly enough, if we look at the “all clean laps” average, the situation bends even more in favor of Toyota:

Porsche #1:      23.292
Porsche #2:      23.304
Toyota #7:        23.2
Toyota #8:        23.243

If we look at the plots relative to the best 20, 50 and 100 first sector times, we can pretty much confirm what the average times told us. On the single quick lap, Porsche seems to have an advantage, but on the all race pace, Toyota is slightly quicker.


LMP1 best 20 sec1


LMP1 best 50 sec1


LMP1 best 100 sec1

Each of the three plots above shows how car n.7 is constantly the quickest in the first sector, excluding the very left portion of the window. Car n.8 is a bit slower and lies below Porsche’s pace for at least the best 5-6 laps, but after the 20-mark becomes clearly the second quickest car.

What about sector 2? The situation looks exactly the opposite as in sector 1, with Porsche cars being clearly and constantly quicker than the two Toyota; no matter if we look at the best time overall, the average of the best 20, 50, 100 or “all clean laps”, Porsche seems to retain a pretty sensible advantage, with car n.1 looking again stronger than car n.2, while the two Toyota are closer to each other.

Second Sector best times:

Porsche #1:      40.476
Porsche #2:      40.786
Toyota #7:        40.815
Toyota #8:        41.122

Second Sector best 20 times average:

Porsche #1:      40.793
Porsche #2:      40.952
Toyota #7:        41.238
Toyota #8:        41.287

Second Sector best 50 times average:

Porsche #1:      40.955
Porsche #2:      41.062
Toyota #7:        41.385
Toyota #8:        41.403

Second Sector best 100 times average:

Porsche #1:      41.161
Porsche #2:      41.210
Toyota #7:        41.576
Toyota #8:        41.571

Second Sector all clean laps average:

Porsche #1:      41.573
Porsche #2:      41.687
Toyota #7:        42.031
Toyota #8:        41.935

The plots of the best 20, 50 and 100 sector times tells more or less the same story. Toyota’s lines are always and sensibly above Porsche’s ones.


LMP1 best 20 sec2


LMP1 best 50 sec2


LMP1 best 100 sec2


Again, we could speculate trying to understand why Toyota is slower than Porsche in this sector. Personally, it is for me not straightforward to come to a conclusion, because sector 2 is actually a combination of a very demanding sequence of corners where both Aerodynamics and mechanical grip play an important role, then followed by two segments where top speed, hybrid boost and corner exit traction are surely very important.
Analyzing top speeds, we can try to get a better insight about the main performance differences between the two manufacturers, assuming both cars has a similar overall engine power and similar boost effects on terminal speed (which is unfortunately not necessary the case, but we don’t have any better data than this to try to understand where the differences are).
It is not hard to notice how, since the introduction of their high downforce kit in Nürburgring, Porsche seems to have gained a good advantage on Toyota, in terms of pure performance. The 919 has most probably more downforce than the TS050, maybe even with a small edge in term of aerodynamic efficiency, but here we are already speculating. What we could see analyzing both marques top speeds is that the Porsche (above all car n.1) seems to be a bit slower than the Toyotas, while car n.2 has more similar performances.
If we look to the plots relative to the best 20, 50 and 100 top speeds achieved during the race, we can identify a pretty clear trend.


LMP1 best 20 TS


LMP1 best 50 TS


LMP1 best 100 TS


The first thing we can clearly see is that Porsche n.1 has constantly the lowest top speed, if we exclude the best top speed overall. This could mean, assuming the coasting strategy were not too different than the one of Porsche n.2,  that car n.1 maybe had an higher downforce set up and, even if paying this in terms of drag, this choice seems to have paid in terms of performance.
The second thing is that Porsche n.2 and Toyota have similar top speeds, but Porsche n.2 is still quicker than both Toyota cars in the second sector. This seems to suggest that the 919 has effectively a similar drag compared to the TS050, but a higher downforce and hence a better efficiency.
We could get some more data to evaluate this point also by looking at the third sector times. Last sector’s performance is also very much handling-driven, since the combination of slow and fast corners surely requires a very good combination of low speed (mechanical) grip, downforce and traction.

Let’s start looking to the best sector time overall and best average sector times first.
The situation is pretty well balanced, with Toyota having a small advantage on Porsche:

Third Sector best times:

Porsche #1:      43.103
Porsche #2:      43.266
Toyota #7:        43.073
Toyota #8:        43.079

The two Toyota are amazingly close to each other and a bit quicker than the two Porsche, with car n.1 following very close and car n.2 being about 0.2 seconds slower.
The situation remains extremely close even when looking to the best 20, 50, 100 and “all clean laps” averages:

Third Sector best 20 times average:

Porsche #1:      43.579
Porsche #2:      43.603
Toyota #7:        43.584
Toyota #8:        43.461

Third Sector best 50 times average:

Porsche #1:      43.841
Porsche #2:      43.795
Toyota #7:        43.727
Toyota #8:        43.726

Third Sector best 100 times average:

Porsche #1:      44.078
Porsche #2:      44.016
Toyota #7:        43.899
Toyota #8:        43.940

Third Sector “all clean laps” times average:

Porsche #1:      44.453
Porsche #2:      44.462
Toyota #7:        44.360
Toyota #8:        44.361

As we said, there is in general a pretty well balanced situation, with one car gaining or loosing its ranking position depending on the average we consider, but with differences to its mate car and to the other brand that stay more or less always pretty constant.
Toyota is slightly quicker in this segment, but the gap between the two brands is really small.
This is confirmed also by the plots relative to the best 20, 50 and 100 sector 3 times.


LMP1 best 20 sec3


LMP1 best 50 sec3


LMP1 best 100 sec3


We can see the Toyotas are constantly slightly quicker than the two Porsches, with car n.8 signing the best 20 laps, but falling a bit behind car 7 after the 25th best lap mark.

A final note about the pit stops each team did. The strategies proposed by Toyota and Porsche were pretty much aligned, with both teams double stinting their tyres. The main difference was that Porsche changed the tyres in the first pit stop and double stinted the second set, while Toyota went for double stinting the first set and did a first shorter pit stop.
Anyway, if we look at the overall time spent in the pit the two teams were pretty much aligned, with Toyota overall pit time being only slightly higher.

Porsche #1:      396.473
Porsche #2:      396.361
Toyota #7:        397.762
Toyota #8:        398.31

Also, all pit stops happened very close to each other.



The LMP2 class has also produced a very interesting race, with exciting track action. Even if the WEC misses the technical variety of the ELMS series in terms of chassis manufacturers, the extremely high quality of teams and drivers ensure always a top level show.

COTA races was dominated by Signatech Alpine, that was able to succeed with a comfortable gap on the second classified car (Rebellion car n.13) even with one more pit stop of about 46 seconds, required to repair a a rear light and even if they had to double stint the tires with Menezes (who felt several positions during that stint), because they used two sets during the qualifying. Also, the safety car that was sent on the track during the race surely didn’t help them, but they really have been stronger of any odd in Texas.

We will look at the first five classified cars, run respectively by Alpine, Rebellion and DC Racing / Jota.

All of them did 7 pit stops during the race, with the exception of the winner crew (already mentioned) and DC Racing Car n.37, that pitted 8 times.
The overall pit stop times are shown here:

Signatech – Alpine #36:       546.293
Valiante – Rebellion #13:    506.293
Valiante – Rebellion #31:    536.899
DC Racing #38:                     482.938
DC Racing #37:                     546.526

In terms of performance, car n.36 has always been among the quickest cars on the track, also running the fastest lap during the race and being clearly the fastest if we look at the 100 best laps and “all clean laps” average:

Best Lap times:

Signatech – Alpine #36:             115.427
Valiante – Rebellion #13:          115.616
Valiante – Rebellion #31:          115.695
DC Racing #38:                           116.094
DC Racing #37:                           115.434

Best 20 lap times average:

Signatech – Alpine #36:            116.578
Valiante – Rebellion #13:         116.532
Valiante – Rebellion #31:         116.925
DC Racing #38:                          116.781
DC Racing #37:                          116.302

Best 50 lap times average:

Signatech – Alpine #36:            117.190
Valiante – Rebellion #13:         117.174
Valiante – Rebellion #31:         117.620
DC Racing #38:                          117.542
DC Racing #37:                          117.246

Best 100 lap times average:

Signatech – Alpine #36:            117.728
Valiante – Rebellion #13:         117.870
Valiante – Rebellion #31:         118.099
DC Racing #38:                          118.284
DC Racing #37:                          118.351

“All clean laps” lap times average:

Signatech – Alpine #36:            118.569
Valiante – Rebellion #13:         118.997
Valiante – Rebellion #31:         118.982
DC Racing #38:                          119.564
DC Racing #37:                          119.698

The first thing to notice is that car n.36 looses the “lead” in the above lap times averages only in the best 20 and 50 lap times average (still being anyway always very close to the fastest car), but is clearly in front of all the others again in the best 100 lap times average.
It is also interesting to notice how car n.37 has indeed a very competitive pace if we look at the best lap times, best 20 and 50 lap times average, being either the fastest or very close to the fastest car in each of these classifications. Anyway, its performance falls down if we look at the best 100 and “all clean laps” lap times average, with a pretty big gap to car n.36.
It is not a case that, during the first half of the race, car n.37, with Brundle at the wheel, was able to climb the ladder up to P1, with a lot of very exciting overtakings. If we look to the best 20 laps plot, Car 37 is extremely competitive.

LMP2 best 20 laps


Anyway, already looking at the best 50 laps plot, the situation changes completely and car n.37 shows a very strong fall off after the 25th mark


LMP2 best 50 laps


Looking at the best 100 laps, finally, we see as car 37 performance drops really dramatically with the crew becoming the slowest of the first 5 classified cars after the .. mark:


LMP2 best 100 laps


On the other side, looking at the best 50 and 100 laps plots, we can see how car 36 is the one staying more constant in terms of performance, not producing the best lap times up to the 10-15 mark, but clearly becoming the quickest after the 35 mark, showing the smallest performance degradation.

It would be interesting to try to understand what exactly lead to car n.37 being so fast in certain phases and much slower in other moments of the race. As we have seen also analyzing ELMS Paul Ricard race, the pace of the non-professional driver seating in the car can really define the race, but was it really a problem with the slower driver here?
If we look at following plot, showing the lap times every car did at every lap, we can see once again, on one side, how constant car n.36 was, but also how other car was affected by both different drivers performance or by performance degradation:


LMP2 stints analysis


Although the plot seems simply messy, we can identify a few interesting points:

  • car n.36 line is the one staying averagely more horizontal
  • same cars had apparently very slow stints, like car n.38 in its second and car n.13 in its third
  • some cars show more performance degradation during a single stint, like car n.37 in the last stint and car n.38 in the third one.
  • car n.37 was the quickest on track during the second stint


If we compare each car to car n.36, it is a bit easier to identify what happens during each stint:

LMP2 stints analysis 36-13.JPG


This first plot shows car n.13, compared to car n.36. We can see how car 13 was particularly slow in its third stint (the non professional driver was in the car), but slightly quicker in stint n.2. Also, both cars seems to show a pretty constant performance in each stint, with the lap times line not going up too aggressively (which means no extreme pace degradation), if we exclude stint 3 for car n.13.

The following plot is a comparison between car n.31 and car n.36:


LMP2 stints analysis 36-31


Both cars seem pretty constant, but car n.31 is slightly slower during pretty much the whole race (excluding maybe the fourth stint), as confirmed also by the best laps plots.


LMP2 stints analysis 36-38


If we look at the previous plot, showing a comparison between car n.36 and car n.38, things get more interesting. Car n.38, even if showing sometimes comparable pace to car 36, seems to suffer of a much stronger performance degradation during a single stint, as we can see looking to the second, fourth , sixth and last stint (lap times lines goes up). We could suspect this happens mainly during the second stint done with a certain set of tyres, but i am honestly not sure this is really the case, since i could not follow when they exactly did their tyre changes.
The same tendency is even more evident if we look at the comparison between car n.37 and car n.36:


LMP2 stints analysis 36-37


Car n.37 has a tremendous pace during its second stint (ad this is well confirmed by the 20 best laps plot, where car n.37 is clearly the fastest), but is slower in pretty much all the other stints and shows also a much more pronounced performance degradation in every second stint, as the sister car n.38.

I will not go through the track sectors analysis for the LMP2 class, since, because of the absence of chassis variety in the WEC, we would simply confirm the same tendencies we saw analyzing the overall lap performance.

Posted by: drracing | September 11, 2017

ELMS – Paul Ricard race analysis

Hi everybody!
This will be a different post than what I normally do in recent times. I will try to dig into something I have never done before, hopefully without coming to completely wrong conclusions.

Before i dig into today’s topic, a short update about the latest activities i have been involved with.
I am happy to say that also in 2017 something i wrote will be featured in the “24 Hour Race Technology” magazine! Because of this article i came into contact with many extremely well prepared people and that also gave a chance to validate the latest iteration of 2017 LMP2 vehicle model i built, now based on another manufacturer data. It was again extremely interesting to see how close simulation results and logged data are to each other!

More about this soon!

Let’s now start with what i wanted to discuss today.

A couple of weeks ago the European Le Mans Series had its fourth race of this season in Paul Ricard, in France and the race offered some food for though.
First of all, the track layout used this year is not the same as in 2016, because the long Mistral straight had been cut in two using one of the chicane options available.
As far as I understood, this decision has been taken to make the track safer, probably thinking about the top speed that the LMP2 cars could achieve at the end of the Mistral straight, to then run into a very, very quick corner. Anyway, in my very humble opinion, this has made the track much less interesting and, “de facto” eliminated probably the only very high speed corner that Le Castellet circuit had to offer (making probably also the decision about how much downforce / drag to run easier, because of the absence of the very long straight).

The track is known to have a very flat and leveled surface, with less bumps compared to other circuits and features now mainly low-to-medium speed corners (first, second and third gear).

The most interesting thing was the first victory of SMP team and Dallara since the new LMP2 rules and cars were presented. The team competes with two very young Russian drivers coming from single seaters and who seem to have found very quickly a good feeling with the LMP2 Dallara. This was their second race in LMP2 and already in Austria, with nearly no practice before the race, they obtained very encouraging results.
With this article I will try to analyze how they came to a win and where they did better than the others. To do this, we will mainly use the data made available by the organizer after each race.

Before we look at numbers and plots, anyway, one thing to consider is that SMP was the only team having only two drivers that, although being classified as Silver, are actually more or less Professionals and are both reasonably quick. As we will see, this plays a very important role in defining a team’s strategy and results.

Beside this, the first thing that we can notice is that the SMP team has been the only one to go through the race with only four pit stops, while all the other teams (who didn’t encounter car issues) did five. The time they spent in the pit during the race was 4:51.592, while the team that ended in second place (G-Drive, Oreca n. 22) spent 6:24.037 minutes in the pit, with a difference between the two teams of about 1 minute and 22 seconds. The gap between the two cars on the finish line was about 1 minute 28 seconds, so it seems that a big part of it was actually up to the one less pit stop that SMP did.
SMP doing only four pit stops came for sure as the result of a specific strategy but also of a gamble, supported very well by a Full Course Yellow that came close to the end of the race, which surely helped to save some fuel. On the other hand, it looked like the Dallara was a tiny bit more gentle on the tires than the Oreca and car 22 (second classified) was forced to change the rear tires at the end of the race (during their latest pit stop) to avoid problematic balance issues, thus losing precious time.

Interestingly enough, anyway, SMP car was surely not the one with the best pace on the track.
If we simply look at the best lap time that each car achieved during the race, the SMP team has a 1.55.756 on his list, while car n. 22 (G-Drive by Dragonspeed, Oreca07) did a 1:55.252 and car n. 21 (Dragonspeed, Oreca07) did a 1:55.463.
Also looking to the average of the twenty best lap times that each car did, we have the following situation:

27 SMP:                      1:56.937
22 G-Drive:                 1:56.717
21 Dragonspeed:         1:56.655

The Oreca still seems to have a small edge on the Dallara on this track, although Paul Ricard is probably one the tracks that suites SMP car and setups better.
The situation anyway bends a bit toward SMP if we look at the average of the best 50 lap times:

27 SMP:                      1:57.672
22 G-Drive:                 1:57.647
21 Dragonspeed:         1:57.439

This shows a substantial parity between SMP and G-Drive and a smaller advantage of Dragonspeed compared to the 20 best laps average.
Finally, if we look at the average of all lap times (considering only clean laps, so no FCY and no pit-in / pit-out laps), the situation bends even further to SMP side:

27 SMP:                      1:58.66
22 G-Drive:                 1:58.958
21 Dragonspeed:         1:59.871

As far as I can see, this shows exactly the advantage of having a crew composed “only” of two quick silver drivers in the car. As we will see shortly, the main disadvantage for at least one of the two Dragonspeed-managed cars was in the time lost when their non-pro driver was at the wheel.

If we look at the following plots, the small advantage of the two Dragonspeed Orecas seems to be well confirmed. The first picture depicts the quickest twenty laps of each car, the second the best fifty.


20 best laps


50 best laps


Looking to the first plot first (quickest twenty laps for each car), it is pretty clear how, on the pure pace, the Oreca is consistently faster than the Dallara. This is especially true with car 21, during the two stints driven by Ben Hanley, but seems to also be the case for car 22, although the difference to the Dallara looks smaller, above all if we consider the right part of plot.
The second plot seems to tell, again, that the car 21 was the quickest, while car 22 and car 27 (SMP Dallara) were much closer, especially on the right side of the 20 mark. This suggests a more constant performance of the Dallara and, maybe, a better tyre management.
It is also interesting to notice how, on the very right side of the plot, the lap times of car 21 jumps up, going above the green line representing SMP car. As we will see, this is what happens when the crew has a non-professional driver unable to drive at a comparable pace than the two other team mates.

It is also interesting to look into the sector times that each car produced. First of all, the image below shows how the track was divided:




Considering sector 1 (which is a mix of straights and second and first gear corners), we can identify a similar situation to what we have seen for the overall lap times. Considering each car overall best sector 1 time, the two Oreca (car 21 and 22) still have an edge on SMP Dallara, with car n.22 being slightly quicker than car n.21 and the Dallara being slightly more than one tenth slower:

27 SMP:                      32.527
22 G-Drive:                 32.406
21 Dragonspeed:         32.416

Anyway, if we consider again the average of the best 20 and 50 lap times of each car, the gap reduces significantly, with the Dallara achieving equivalent results to car 22 on the best 20 laps average and jumping to second place if we look at the best 50 laps.

Best 20 laps average:
27 SMP:                      32.845
22 G-Drive:                 32.846
21 Dragonspeed:         32.67

Best 50 laps average:
27 SMP:                      33.027
22 G-Drive:                 33.103
21 Dragonspeed:         32.929

If we consider all the clean laps done during the race, SMP has the best average overall also in the first sector, which underlines once again the positive effects of their crew mix (with regard to this point, the Dragonspeed Oreca n.21 is clearly the car in the weakest position, becoming the slowest of the three even if it still was the quickest considering both the best 20 and 50 laps averages):

27 SMP:                      33.348
22 G-Drive:                 33.521
21 Dragonspeed:         33.685

Looking at the plots relative to the best 20 and 50 first sector times, we can see, once again, how the n.21 Oreca was the quickest car, while there is pretty much equivalence between car 22 and 27 on the 20 laps plot and a clear advantage of car 27 over car 22 on the 50 laps one, with car 22 times increasing sensibly after the 15-20 mark, compared to car 27 ones.

20 best sect1


50 best sect1

Sector 2, being practically composed only by a straight and a chicane, is a good indicator about car overall top speed (and, hence, drag) and inline accelerations performance (both braking and accelerating) and shows pretty much equity between the Dallara and the two Orecas.
Car 21 is still the quickest, if we only look at the best sector 2 time overall, but with less than 0.1 seconds difference to car 27:

27 SMP:                      34.279
22 G-Drive:                 34.347
21 Dragonspeed:         34.2

Anyway, if we consider the average of 20 best first sector times for each car, all cars have pretty much the same performance:

27 SMP:                      34.52
22 G-Drive:                 34.56
21 Dragonspeed:         34.51

Also looking at the average of the best 50 laps, the situation doesn’t change significantly, although we can start to identify the same tendency already seen before, with the Dallara coming to the front:

27 SMP:                      34.648
22 G-Drive:                 34.709
21 Dragonspeed:         34.665

Once again, SMP has indeed the best average if we consider all the clean laps done during the race:

27 SMP:                      34.866
22 G-Drive:                 35.023
21 Dragonspeed:         35.176

As for the whole lap and for sector one, it is interesting to notice as, also for sector two, the Dallara seem to be more constant in its performance compared to cars 22 and 21, with less difference between the best lap overall, the best 20, 50 laps averages and the “all clean laps” average. This surely suggests how good and well balanced the crew was, as we said already, but also maybe a better tyre management.

All of this is well reflected looking at the plots of the best 20 and 50 sector 2 times.


20 best sect2


50 best sect2


Both plots clearly show how, on the single lap performance, Car 21 is still clearly the fastest.
Anyway it is also clear to see how, on the long distance, car 21 and car 27 are very close to each other, with car 22 being a bit slower, with the gap increasing as the lap counter goes forward. It is also interesting to see how, after the 43 mark in the second plot, car 21 sector times jump upward leaving car 27 to be clearly the quickest on track.
This suggests two reflections: as we will see also looking at sector 3 times, maybe car 22 and car 21 were using slightly different aerodynamics settings, although belonging to the same team. In particular, it looks like car 21 went for a setup favoring top speed and, maybe sacrificing a bit of downforce, while car 22 probably went for more downforce and, hence, more drag.
The second point regards, again, the winning crew: on one side, we can see that sector two was probably the one fitting SMP car the best, also if we compare Dallara’s and Oreca’s performance; on the other side, it is interesting to see, once again, how close to each other are the best sector times that car 27 was able to produce, compared to that of car 21 and 22, with car 27 probably suffering of less degradation and using the good drivers performance at the best level possible (including traffic management).

The Dallara showing less difference between the best sector time and the three averages we consider, seems to be also well reflected by Sector 3 analysis.
Anyway, being sector 3 the most twisty one, with the only medium-speed corner of the track (turn 12), it seems to be the best one to evaluate each car handling and downforce. It is probably not a case that, in this sector, performance gap between Oreca and Dallara is bigger, also considering the 50 laps and “all clean laps” average. This is even more interesting, if we keep in mind how car 22 in particular seemed to suffer of a bad tyre degradation during the last stints.

Let’s look again at the best sector 3 time overall for each car:

27 SMP:                      48.711
22 G-Drive:                 48.391
21 Dragonspeed:         48.328

Car 21 is again the quickest, with nearly 0.4 seconds gap to the Dallara and only a small difference to car 22.
Looking at the 20 best sector 3 times average, Car 21 is not the quickest though, with car 22 now jumping on top and the Dallara still being about 0.4 seconds away:

27 SMP:                      49.439
22 G-Drive:                 49.085
21 Dragonspeed:         49.183

The best 50 laps average shows a similar situation, with a the gap between the two marques closing a bit:

27 SMP:                      49.751
22 G-Drive:                 49.538
21 Dragonspeed:         49.631

The interesting part about this sector comes anyway when looking at the “all clean laps” average. For the first time SMP is not on top and the n.22 Orecas still has an edge (again, even with the car suffering of bad tyre degradation at the end of the race):

27 SMP:                      50.246
22 G-Drive:                 50.187
21 Dragonspeed:         50.599

Once again Car 21 is affected by the Bronze driver, while car 22 remains on top.
We immediately notice that, what we identified looking at sector 2 times could be true, with car 22 being, on the longer distance, a bit quicker than car 21 in this third sector, probably because of different setup choices. Even more interesting is how, also on the “all clean laps” average, the Dallara remains slower than the Orecas (inverting a tendency that could be saw in both overall lap times and the first two sectors times), in a track section where handling and downforce probably pay a very important role.

This is confirmed also looking 20 and 50 best sector 3 times plots.


20 best sect3


50 best sect3


We again see how SMP crew is constantly slower than G-Force and Dragonspeed ones, if we only exclude car 21 last 10-15 quickest laps in the 50 best sector 3 times plot.
Equally interesting, if we just exclude the best sector time overall, car 22 pace looks always stronger than car 21 in this sector, with better times even when looking at the 50 best times plot and at the very right side of the graph and this underlines again how car 21 and 22 could have probably gone on two different setup strategies.

Finally, let’s look shortly at the lap times history during the whole race for each car. The following plots show the lap time vs lap number for the whole race. Vertical peaks indicate either a pit stop or a safety car situation.


All Race Laps


This plot only looks at cars 21, 22 and 27. The first thing to notice is that, as we already mentioned several times, car 21 significantly suffer a pretty slow driver in the crew, compared to Ben Hanley and Nico Lapierre. Every time the third driver is in the car, lap times increase by about three seconds: this happens namely between lap 22 and 47 and again between lap 68 and 87.
The most constant car, in term of lap times, is clearly the SMP Dallara and that seems to pay off well above all at the end of the race, when the car in second place (the G-Drive Oreca, n.22) suffers of a sensible performance reduction.
This plot also underlines once again how, if we look at pure performance, car 21 is clearly on top, with car 22 following closely. Still, Car 27 looks extremely consistent during the whole race and, also because of a better strategy (including one less pit stop), it managed to win with a pretty good margin.

Posted by: drracing | June 16, 2017

Testing suspension geometry in the simulator

Hi everybody!

As anticipated in my last post, this article will be a review of one of the project I was involved with during the winter. As for one of my latest project of 2016, also this case study is actually relative to an investigation I supported using driving simulation as a development tool, in a way I had not thought possible a few years ago. Life is full of surprises.

After during the summer I had a chance to prove myself how, even using a cheap software, a proper vehicle model could be used effectively to gain some more insight about car setup and its intricacies (helping the team I was supporting to further improve a bit their performance), this time I was asked to try to evaluate different design solutions and quantify not only their performance but also driver’s perception / feedback to each of them.
The study was focused more specifically on investigating the effects on car handling and performance of different front suspension layouts, differing sometimes pretty much one to the other (in one or more areas), with the aim to evaluated what the driver would feel driving each of them, together with their impact on the car-driver system’s performance.
The guy who asked me to support this project was initially mainly looking to understand the impact on driver steering feedback coming from different designs / geometries, but soon the study also evolved into a chance to evaluate which influence each setup would have on overall performance and why, identifying, also through data analysis, in which areas there would be the biggest differences and the reasons behind each Delta.

The model we used was actually the reliable and good validated 2016 LMP2 one, driven in Silverstone.
We starting establishing a baseline (lap time / performance and behavior) using the original design, then moving onto testing new solutions.
The parameters we worked on at the beginning were mainly scrub radius, king pin angle and caster trail, but inevitably we soon moved on also acting on camber change, roll center position and migration and caster angle and finding out how also other parameters that we underestimated at the beginning actually play an important role, at least in terms of steering feedback.

Before I dive a bit more into the details of this project, let me spend two words about the assumptions used by rFactor in terms of suspension modeling and my view about the software’s limitation.
First of all, it is clear to me, as it is probably clear to all of view, that rFactor was not though as an engineering tool and there are some limitations that are difficult to completely overcome, if not acting on the code itself (which is out of my skills, intention and interest).
I would of course be very happy to have a chance to use more “engineering” oriented and flexible software for my projects, but the budget required to buy one of them would probably be absolutely out of reach even for many small companies.
The truth is, anyway, that it is amazing how much you can test and understand, in terms of vehicle dynamics and even in terms of setup / layout even using a cheap product like rFactor, despite all of its limitations. Of course, “conditio sine qua non” is always to know exactly your assumptions and to build a model as close as possible to the real car (or at least to the data we have about it).

As we had a chance to briefly mention in one of my previous posts, rFactor simulates suspension behavior in a very advanced way, since it practically define each suspension component as a rigid body (with or without mass properties) connected to the surrounding ones through mechanical constraints, like spherical joints, hinges, etc. This means, for example, that the wheel and the “spindle” (which actually identifies the complete Upright – Hub assembly in rFactor) have their own (definable) mass and inertial properties. The links are defined as rigid elements with no mass, practically locking the distance between two points depending on their length.
There are a number of things to take care of, when modeling a suspension, including like rFactor adjust camber, caster and toe. But if these parameters remained locked or the “errors” that the software does when changing one of them are compensated in a proper way, the suspension kinematics is simulated in a very similar way to what a normal multibody package would do.
A bit of attention must be paid for layouts using a pushrod / pullrod with a rocker to activate the damper/spring unit, since the Rocker assembly and its functionality cannot be reproduced. Of course, since we are dealing with an LMP car, this is exactly the situation we find ourselves in, as we are considering a double wishbone with pushrod actuated rockers for both front and rear axle.
From a wheel rate / motion ratio perspective, this “issue” can be easily overcome producing very accurate results, as we have seen already in the past, the point here simply being to directly work with the Wheel Rates, instead of the spring rates. To do this, I am using a trick, playing with the pushrod’s hardpoints (and, consequently, with its orientation), to obtain exactly the wheel rates and wheel rates change (with respect to wheel travel) I want, independently from the rest of the kinematics.
This means that the “virtual pushrod” will have another position and another 3D orientation compared to the real one, in order to obtain the same wheel rate and wheel rate change of the real car.

From a statics perspective and, more specifically, if we want to isolate and quantify the load acting in each linkage (this is particularly important for this study, since the steering feedback was one of the parameters we would like to better understand and since rFactor and the plugins I used produce the steering wheel feedback strictly basing on the force acting on the steering tie rods), this actually can bring a difference in the results because, depending on the position and orientation of each beam, the loads acting on each of them (for a defined loadcase at the contact patch) will change. In other words, although the suspension “as a system” would still behave the same (at least considering every suspension link and the upright-hub assembly as a rigid body, so ignoring every compliance), the internal reaction forces would change.
Anyway, the good news here is that, for the suspension geometry and layout we are using, even if the absolute value of the force acting on the steering rod that could be measured in the real car is different than the one produced with our “rfactor alternative-pushrod position”, its trend and gradients are very similar: in other words, the load acting on the steering rod is changing in a similar way in both cases, for a certain change of the Fy, Fx or Fz acting at the contact patch. All of this has been checked first using a simple excel sheet I built to calculate the reaction loads in each suspension members, depending on the loadcase at the contact patch and on the pickup points position; to be absolutely sure, a second check has been done using a Multibody software and in both cases the results have shown exactly the behavior I described above.
What is important for this study is actually the trend shown by the reaction forces in the steering tie rod, since the steering forces produced by the simulator’s steering system are proportional to the steering tie rod loads through a factor that is actually what we use to tune the intensity of the steering feedback itself.

Beside the driving simulation itself, I also used some CAD and some Multibody modeling for this project; on one side, I did this to evaluate each layout before to test it with the simulator but, most often, also to tune each geometry a bit and try to isolate and change only the parameters we wanted to evaluate at each step, keeping the others as constant as possible, with the aim to limit their influence on the results (in particular the steering torques, but not only).
Anyway, in certain cases this was not actually possible because of real world physical limitations (for example parts colliding to each other) and we had to accept that, with a certain setup, some parameters that we would not want to change would indeed change (see for example Ackermann effect). It was nonetheless interesting to see how a certain set of parameters would work, both in terms of performance and driver perception.

It is probably not a case that the best performing solution was also the one which was received more enthusiastically by the driver (although, in my experience, this is unfortunately not always the case).

The first test was done using the baseline configuration, namely the front suspension geometry originally built in the car. We run some sessions with this solution and established a reference lap (and the relative logged data) to be used later as a base to evaluate the following setups both subjectively and through data analysis.

The second test was performed using a revised front suspension geometry (we will call it Exp1), mainly deviating from the baseline in the upright area. In particular, caster angle was reduced by slightly more than one degree, although the caster trail was slightly increased. The King Pin Angle was also increased by about 2 degrees, producing a smaller scrub radius. This setup also had a slightly higher Ackermann effect (more dynamic toe out) and a slightly higher roll center.
The feeling was immediately very good. The car showed a bit more understeer, above all in mid-corner and, sometimes, even in corner exit. This helps in certain situations, see for example turn 3 (a 1st gear corner, where traction is very important), but seems to be a limit in others, see for example the last chicane, above all in the exit of the second (right) corner leading to the last right tender before the box straight.
Beside showing more mid corner understeer, this setup actually also produced a more reactive behavior in corner entry, above all in quick corners.
From a driver feedback perspective, it was interesting to notice how the changes we did made the car easier to drive, giving to the driver some more confidence and also the feeling that the relationship between steering angle and front cornering forces remains more or less linear for the complete range of used angles (while the baseline had a more unpredictable behavior, above all at very high steering angles, where the cornering force seemed to drop more abruptly and in a less predictable way when exceeding with the steering angle and the driver had a feeling of “loosing front grip”), although producing more understeer.

An interesting point was the increase of steering forces, perceived by the driver and confirmed by data analysis. This seems to go somehow against the expectations of some the changes we did (see for example a reduction of caster angle).
Anyway, two factors must be considered: on one side, the caster (or longitudinal) trail was slightly longer (granting a longer lever arm to Cornering forces in producing Self aligning torques); on the other side, we noticed how, one of the side effects of the geometry change we did was a slight reduction of the length of the Steering Arm (distance between the outer tie rod point and the steering axis), thus creating a bigger force in the steering tie rod for a certain torque to be reacted.

In terms of performances, this setup produced a 3-4 tenths improvement compared to the baseline.

Here below you can see the traces relative to three logged parameters: speed, steering angle and steering force. The tested geometry is shown in red (Exp1).

Exp1 - speed

Exp1 - steering

Exp1 - steer forces

The third test (referred here as Exp2) was performed using a geometry differing from the baseline much more aggressively. Caster has been increased by about 3.5 degrees, producing also a substantially bigger caster trail (about 30% bigger than the baseline one). King Pin angle and Scrub radius were very similar to the original configuration. The Ackermann effect (or percentage or dynamic toe) was slightly lower, as was the camber gain. Finally the static roll center position was also substantially lower.

Driver’s feedback was again very good, although the car behaved very differently than in the previous test. A strong reduction of the understeering tendency was evident from the very first corners, with the front axle now having substantially more grip, above in mid-corner but also in corner exit, with a reduction of the power understeering tendency that disturbed a bit in the last chicane in the previous test. Interestingly, even showing much more front grip, the car didn’t become instable or unpredictable, nor it showed any “dangerous” oversteering behavior.
In quick corners, the vehicle was now a bit less reactive than in the previous test, but still showed very good driveability.
Steering forces increased slightly, compared to the previous test and were thus sensibly higher than the baseline setup.

The lap times we obtained were anyway very similar to the previous test, with this latest outing producing an about 0.02 seconds quicker lap time.

In the following plots, this test’s results are shown in orange (Exp2).


Exp2 - speed

Exp2 - steering

Exp2 - steer forces


The direct comparison between this setup and the previous one (not shown in these pictures) show slightly higher steering forces in this latest case, although the steering arm being now bigger than both the baseline and the first test.
To finally quantify the influence of this latest parameter we later did a very quick test using a very similar geometry compared to the one tested here, but reducing the steering arm to a very similar value compared to the baseline (and trying to keep the other parameters the same). This further increased the steering forces, making them substantially higher than both baseline and first test and, more important being a direct comparison, much bigger than this test 2.

I will not bore you with a description of all the other tests we did (some of them were not as successful as the first two and the following one) and I will just jump to the one that produced the best performances and the best driver feeling, also to show how much of a difference it did in terms of lap times and how good this could be felt in the simulator.
This latest test was performed using a front suspension geometry now having about 0.8 degrees more caster than in the previous one (and so about 4.3 degrees more than the baseline), a slightly bigger caster trail than test 2 (but much bigger than the baseline) and a smaller scrub radius (compared to both baseline and test 2), because of a sensibly bigger king pin angle. Roll center, camber gain, Ackermann effect and steering arm’s length were all kept practically the same as the baseline.
One of the interesting features of this geometry was how the camber evolves with the steering angle, not only on the outside tire but also on the inside. A caster increase normally always produce an increase of camber delta as a function of steering angle, but here this phenomenon seems to be amplified and probably also the contribution of an “effective” camber change on the inside tire helped.

Dirver’s feedback was immediately enthusiastic, as also proved by the lap times, that dropped by about 3 tenths of a second compared to the previous two tests (so overall about 7 tenths compared to the baseline).
The driver felt now even less understeer and the front axle, which ensures a very high grip, without generating oversteer or instability in any corner. The general feeling was actually of more grip on both the front and rear axle, but with even less understeer than the previous test.
This less understeering tendency could be felt also in corner exit, but somehow that didn’t deteriorate traction. The car was much easier to drive and allowed to push more easily, communicating always a feeling of stability and predictability. As a result, it was possible to go easier on the throttle on some corners’ exit, like the last chicane.
This was also true for the steering feedback: steering forces were now higher than any of the previous tests and the “linear” feeling we described about the first test remained, with the driver reporting that, even with very high steering angles, the front axle always seemed to produce cornering forces in a gradual manner, without abruptly loosing grip or generating oversteer or instability.
Interestingly, the steering trace data seems not always to confirms this strong grip of the front axle, at least compared to the baseline logged data. But it is also true that it still shows a more homogeneous use of the steering wheel, probably confirming how this more linear tendency in the relationship between steered angle and front grip and allowing the driver to maneuver the car more aggressively.

All of this seems is shown in the following plots relative to this test’s logged data, depicted in pink (Exp4).


Exp3 - speed

Exp3 - steering

Exp3 - steer forces


Beside finding extremely interesting all the tests and their outcome, I was once again amazed to see how much we can learn using (properly and also knowing the limitations and the turnaround to be used to compensate for some simplifications) a simple and cheap tool like rFactor, even from an engineering perspective.
The goodness of the results of this study is of course strongly dependent on how good also the vehicle model was built, in particular for all what concerns the tire model. I am pretty sure my model adheres pretty good to the data I have, but of course any model is only a model. Real testing is and will always be the best way to validate and decide on certain design decisions, but it is amazing to experience how much we can do in terms of pre-evaluation even with such a cheap tool.

By the way, it was really good fun to go through all these designs and also have a chance to see how they perform, getting a “more real” feeling of how the car would handle and being able to do 1:1 tests, with all the advantages connected to the use of a simulation (see deletion of the effects of track conditions, temperatures, rubbering, traffic etc).

« Newer Posts - Older Posts »